Paperlane

no. 10 · from the archive← Thu 5 Nov archive Mon 9 Nov →

Friday, 6 November 2026

652 papers in 32 fields

Combinatorics50

For every overlap $m\geq 3$ and every pattern length $K\geq 2m+1$, we construct an explicit permutation $p\in\mathfrak{S}_K$ such that the number of permutations containing $j$ prescribed occurrences of $p$, with successive starts separated by $K-m$, is $(2j-1)!!$ for every $j\geq 1$. Additional occurrences are permitted. We prove that the corresponding occurrence poset is the ordinal sum of a chain on $m+j(K-m-2)$ elements and a comb on $2j$ elements. An insertion bijection counts the linear extensions of the comb. This gives the uniform representatives requested in the double-factorial part of Question 6 in Section 7 of Blitvić, Kammoun and Steingrímsson, arXiv:2610.29846v1. A Lean 4 formalization and its recorded verification data accompany the paper.

comments
5 pages. Lean 4 formalization and verification records included as ancillary files
fields
Combinatorics
pdf ↗ page arXiv:2611.06937 ↗

For a finite symmetric nonnegative dissimilarity matrix $D=(d_{ij})$ with zero diagonal, we study the high-resolution limit ($t\to \infty$) of $\mathbf{1}^T(e^{-td_{ij}})^+\mathbf{1}$. We prove that this limit always exists in the extended real line and is rational whenever finite, without weightability or invertibility assumptions. At fixed cardinality, a finite rational polyhedral fan governs rank and leading asymptotics. Consequently, only finitely many finite limits occur, and every limit admits an integer-data realization at the same cardinality. In contrast, allowing cardinality to vary, every rational pair of zero-pattern magnitude and high-resolution limit is realized using dissimilarities in $\{0,1,2\}$; adding the level $3$ permits either infinite limit with any rational zero-pattern magnitude. We also classify all limits on at most four points. Moreover, we derive an exact Schur decomposition valid under persistent singularity. Finally, it is shown that positive semidefiniteness at some positive scale is equivalent to equality of dissimilarity rows for every zero-distance pair, in which case the limit counts the zero-distance classes.

comments
29 pages
fields
Combinatorics
pdf ↗ page arXiv:2611.06960 ↗

An $r$-uniform hypergraph $H$ is uniformly dense with respect to $j$-graphs if, up to a small error, at least a prescribed proportion of the $r$-cliques of every $j$-uniform hypergraph on the same vertex set are edges of $H$. For a forbidden hypergraph $F$, its uniform Turán density at level $j$ is the supremum of these density bounds over arbitrarily large $F$-free hypergraphs with the error tending to zero. For every $r\ge4$ and $1\le j\le r-3$, this paper proves that this Turán density is zero precisely when $F$ admits a vertex ordering in which every $(j+1)$-set shared by two edges occupies the same positions in both edges. This generalizes the zero-density characterizations of Erdős at $j=0$ and of Reiher, Rödl, and Schacht at $j=r-2$, and establishes the latter authors' proposed extension to the remaining nontrivial levels. Consequently, at level $j$, every positive uniform Turán density of a possibly infinite forbidden family is at least $N^{-N}$, where $N=\binom r{j+1}$. The proof combines hypergraph regularity with adjacent exchanges in spreadable random hypergraphs.

fields
Combinatorics
pdf ↗ page arXiv:2611.06964 ↗

A conjecture of Sheehan states that every Hamiltonian $4$-regular graph contains a second Hamilton cycle. A natural problem is to find the smallest positive integer $k$ such that every Hamiltonian $r$-regular graph contains a second Hamilton cycle when $r\geq k$. Thomassen proved that $k\leq 300$. Improving Thomassen's result, Haxell, Seamone, and Verstra{ë}te proved that $k\leq 23$. We improve this bound by proving that every $r$-regular Hamiltonian graph contains a second Hamilton cycle whenever $r\geq 17$. Our new idea is to establish a transversal theorem for set systems on a cycle using Harris's results on the Moser–Tardos algorithm for Local Lemma. We then apply Thomassen's theorem on independent dominating sets to obtain a second Hamilton cycle. For $r\ge 18$, we also give some algorithms to find such a transversal.

comments
19 pages
fields
Combinatorics
pdf ↗ page arXiv:2611.06985 ↗

We show that every $K_4$-free graph on $n$ vertices has a balanced bipartition such that neither part contains more than $0.0671n^2 + o(n^2)$ edges. This improves the previous best bound of $0.074n^2$ by Xu and Zhang. We also formulate a conjecture that the tight bound for this problem is $\frac{425}{6498}n^2 \approx 0.0654n^2$, and present a construction showing that it would be best possible.

fields
Combinatorics
pdf ↗ page arXiv:2611.07082 ↗

We establish a new asymptotic lower bound for the volume of a maximal partial Room square of even order $n$. We prove that its volume is at least $\frac{2\sqrt{66}-6}{57}n^2-O(n)$, improving the leading coefficient $1/8$ in the lower bound obtained by Meszka and Rosa. The proof combines two counting arguments, one based on a transversal of empty cells and the other on the graph of occurring symbol pairs.

comments
10 pages, 1 figure
fields
Combinatorics
pdf ↗ page arXiv:2611.07117 ↗

Let $I_{n,n}$ be a fixed perfect matching of the complete bipartite graph $K_{n,n}$, and let $L_{n,1}=K_{n,n}-I_{n,n}$. A perfect set of $L_{n,1}$ is a $1$-factorization of $L_{n,1}$, that is, a set of $n-1$ pairwise edge-disjoint perfect matchings whose union is $E(L_{n,1})$; a perfect partition of $L_{n,1}$ is a partition of the set of all perfect matchings of $L_{n,1}$ into perfect sets. Brualdi, Chiang, and Li (2005) asked whether $L_{n,1}$ admits a perfect partition. Identifying both sides of the bipartition with $[n]=\{1,2,\ldots ,n\}$, the perfect matchings of $L_{n,1}$ are exactly the derangements of $[n]$, the perfect sets correspond bijectively to the reduced Latin squares of order $n$, and the perfect partitions are exactly the perfect matchings of the $(n-1)$-uniform hypergraph whose vertices are the derangements of $[n]$ and whose edges are the perfect sets. We first give, for every $n$, a constructive lower bound on the maximum number of pairwise disjoint perfect sets of $L_{n,1}$. We then show that $L_{5,1}$ has exactly $42$ distinct perfect partitions and that $L_{6,1}$ has at least $82$ distinct perfect partitions; these are the first examples demonstrating that perfect partitions need not be unique. Finally, we exhibit two distinct perfect partitions of $L_{7,1}$, resolving the $n=7$ open problem raised by Brualdi et al.

fields
Combinatorics
pdf ↗ page arXiv:2611.07118 ↗

We construct an extremal unimodular lattice of rank 37, resolving the last undetermined optimal minimum in ranks at most 48. The construction begins with the Cayley table of $\mathbb{Z}/13\mathbb{Z}$, assigning a vector of norm four to each cell. We prove that the resulting incidence lattice has exactly 26 unimodular completions of minimum four, all isometric. The resulting Cayley lattice $Λ$ has automorphism group $\mathrm{AGL}(1,13)\times C_6$ and is $s$-extremal, with shadow minimum $13/4$ and kissing number $35\,594$. An explicit $4$-frame in $Λ$ yields a Type I self-dual $\mathbb{Z}/4\mathbb{Z}$-code of length 37 and minimum Euclidean weight 16, resolving the corresponding open case in coding theory.

fields
Combinatorics, Number Theory
pdf ↗ page arXiv:2611.07138 ↗

A conjecture attributed to Smith states that in every $k$-connected graph, every pair of longest cycles intersects in at least $k$ vertices. We prove that every two longest cycles in a $k$-connected graph on $n$ vertices intersect in at least $\min\{k,10k-n-30\}$ vertices. Consequently, Smith's conjecture holds whenever $k\ge \frac{n+30}{9}$. This improves the previous bound of $\frac{n+16}{7}$ for $n>33$ and also implies a bound of $\frac{n+22}{9}$ for paths instead of cycles.

fields
Combinatorics
pdf ↗ page arXiv:2611.07145 ↗

An occupied seat in a binary row is called packed if both of its immediate neighbors are occupied. We give a self-contained combinatorial account of this statistic, equivalently the number of overlapping occurrences of 111. An occupied-block decomposition yields a coefficient formula at fixed occupancy and a direct proof of the sharp minimum max{0,3m-2n-2} for m passengers in n seats, including arrangements attaining the bound. The same decomposition gives a rational generating function, recurrences, expectations, and avoidance asymptotics. Full rows, in which every empty seat would become packed upon being filled, are related to independent sets of a path. A neighborhood-union expansion also gives exact counts on rectangular grids. The emphasis is a unified elementary presentation and explicit extremal constructions, rather than a new general enumeration method.

fields
Combinatorics
pdf ↗ page arXiv:2611.07204 ↗

Given a bipartite graph $G$, we denote by $g(G)$ the largest product $|A||B|$ over pairs of vertex sets $A$ and $B$, one on each side, with no edge between them. These pairs are the bicliques of the bipartite complement of $G$; hence, computing $g(G)$ is the maximum-edge biclique problem, which is NP-hard, as shown by Peeters in 2003. Gillis and Glineur characterized it by a continuous quadratic program, in which a large enough penalty on the edges of $G$ makes the global maximizers correspond to the maximum pairs and the local maximizers to the maximal ones. We propose a new threshold for the penalty, which depends only on $g(G)$. For any penalty at or above it, the global maximizers correspond exactly to the maximum pairs, and every local maximizer with positive value corresponds to a maximal pair. This threshold is never larger than theirs and is sharp up to an additive $1/2$. Spectral and semidefinite bounds give computable penalties; in experiments, a simple local ascent succeeds far more often with the semidefinite one than with the earlier thresholds, although on our graphs that bound already reveals $g(G)$. The penalty formulation, whose optimal value is $\sqrt{g(G)}/2$, is also a copositive formulation of $g(G)$. Strengthening a semidefinite bound of Laurent, Polak and Vargas with the completely positive cone gives exactly $\sqrt{g(G)}/2$, while the doubly nonnegative strengthening is already inexact on six vertices. Finally, Pólya's linear programming approximations of the copositive cone become finite at an explicit level and converge to $\sqrt{g(G)}/2$ without ever being exact.

fields
Combinatorics, Optimization and Control
pdf ↗ page arXiv:2611.07239 ↗

A digraph is $s$-quasi-transitive if, for every path $(x_0,x_1,\ldots,x_s)$ of length $s$, either $(x_0,x_s)\in AΓ$ or $(x_s,x_0)\in AΓ$ (or both). In this paper, we give a characterization of $s$-quasi-transitive commutative weakly distance-regular digraphs, where $s=4$, or $s\geq 5$ and the cycle polar diameter is at least $s$.

fields
Combinatorics
pdf ↗ page arXiv:2611.07251 ↗

A quadratic source is a random variable $X\sim\mathbb{F}_2^n$ generated by applying $n$ quadratics to a hidden uniform string. Previously, the best-known explicit extractor for this model required min-entropy roughly $k=n-\sqrt{\log n}$. In this paper, we give an explicit construction for min-entropy $k=\mathrm{polylog}(n)$. Even for the special case of $2$-local sources, where each bit of $X$ depends on two hidden uniform bits, achieving an entropy requirement below $k=\sqrt{n}$ was an important open problem. This is because such sources provably cannot be reduced (in a black-box manner) to our most powerful general-purpose seedless extractor, known as a sumset extractor. Our new extractor bypasses this barrier, providing the first explicit construction to go beyond sumset extraction.

fields
Combinatorics
pdf ↗ page arXiv:2611.07266 ↗

For a parking function, the resulting sequence after all cars have parked is called the associated parking permutation. For each parking function, cars that park in their preferred spot are called lucky cars. In this paper, we enumerate parking permutations that avoid a specified set of permutation patterns in $\mathfrak{S}_3$ and whose lucky cars are the first $k$ cars. Often, the enumeration simplifies to a product of a fraction with a binomial coefficient, recovering existing sequences in the OEIS. For two of the classes, the enumerations are new sequences.

comments
24 pages
fields
Combinatorics
pdf ↗ page arXiv:2611.07267 ↗

For a fixed $k$-uniform hypergraph $F$, let $\mathrm{ex}_k(n,F)$ denote the maximum number of edges in an $F$-free $k$-uniform hypergraph on $n$ vertices. We prove that for every integer $k\ge2$ and every rational $r\in[0,1]$ there exists a $k$-uniform hypergraph $F_{k,r}$ such that \begin{equation*} \mathrm{ex}_k(n,F_{k,r})=Θ(n^{k-r}). \end{equation*} The case $k=2$ is the rational exponents conjecture of Erdős and Simonovits \cite{Erdos1988}, also known as Erdős problem \#571 \cite{BloomErdoswebsite}; a proof was announced in September 2026 and is expounded by Conlon et al.\ \cite{Conlon-Janzer-Jiang-Kim-Lee-Liu-Longbrake-Yepremyan}. For $k\ge3$ the theorem extends their result to $k$-uniform hypergraphs. In the single-hypergraph setting, this strengthens Fitch's finite-family theorem \cite{Fitch} by realizing each exponent with a single forbidden $k$-uniform hypergraph for $0\le r\le1$. We also show, via an explicit two-edge construction, that for $k\ge4$ this strengthening fails for $\lceil k/2\rceil<r\le k-1$: no single $k$-uniform hypergraph $F$ then satisfies $\mathrm{ex}_k(n,F)=Θ(n^{k-r})$. The main result of this paper has been formalized in Lean 4, conditional on two explicitly stated axioms.

fields
Combinatorics
pdf ↗ page arXiv:2611.07290 ↗

Nathanson observed that, for four-element sets $A$ of integers, the most frequent sizes of the iterated sumset $hA$ are separated by triangular numbers. Senger proved this and suggested an analogue for larger sets. We establish this pattern for $k$-element subsets of $\{1,\ldots,q\}$, for every fixed $k\geq3$, where $h\geq2$ is either fixed or grows slower than $q^{1/(2k-1)}$. As $q\to\infty$, almost every such set has the largest possible $h$-fold sumset, and almost every exception has one of $h-1$ explicit smaller sizes. The differences between consecutive sizes among these $h$ values are consecutive $(k-2)$-dimensional simplex numbers.

comments
19 pages. Lean 4 formalization included as ancillary files
fields
Combinatorics, Number Theory
pdf ↗ page arXiv:2611.07303 ↗

A $k$-quasiplanar graph is a graph drawn in the plane with no $k$ pairwise crossing edges. Pach, Shahrokhi, and Szegedy conjectured that every such graph on $n$ vertices has $O_k(n)$ edges. ChatGPT produced a proof of this conjecture. Here, we present a simplified, self-contained version of that proof.

fields
Combinatorics
pdf ↗ page arXiv:2611.07314 ↗

Schubert structure constants are intersection numbers of Schubert varieties. Finding positive combinatorial formulas for these constants remains a major open problem. We resolve this problem when one of the indexing permutations corresponds to a two-row partition. Our approach uses increasing chains, introduced by Sottile for the one-row case, and builds on the skew Schubert functions studied by Bergeron and Sottile. The key step is to connect increasing chains and these functions with the parallel classes of positroid varieties.

fields
Combinatorics, Algebraic Geometry
pdf ↗ page arXiv:2611.07351 ↗

Given a finite Coxeter group $W$ and an element $q$ of a commutative ring $\mathbf{k}$, we define the element $\mathbf{L}_{q} :=\sum_{w\in W}q^{\ell\left( w\right) }w$ of its group algebra $\mathbf{k}\left[ W\right] $. We say that an element $w\in W$ is $W$\emph{-length-symmetric} if its coefficient in the commutator $\left[ \mathbf{L}_{a},\mathbf{L}_{b}\right] $ is $0$ for all $\mathbf{k}$ and all $a,b\in\mathbf{k}$. We prove several sufficient criteria for elements to be $W$-length-symmetric. In particular, we show that any separable permutation (i.e., permutation avoiding the patterns $2413$ and $3142$) in the symmetric group $S_{n}$ is $S_{n}$-length-symmetric. More generally, we construct a notion of separable elements of a finite Coxeter group $W$ (which, for finite Weyl groups, agrees with the eponymous notion introduced by Gaetz and Gao), and show that all such elements are $W$-length-symmetric. This is only a sufficient condition; other $W$-length-symmetric elements include (twisted) involutions and elements of dihedral parabolic subgroups.

comments
39 pages. Comments are welcome!
fields
Combinatorics
pdf ↗ page arXiv:2611.07365 ↗

The Higman-Sims graph is the unique strongly regular graph with parameters $(100,22,0,6)$. In this paper, we show that the only distance-regular antipodal double cover of the Higman-Sims graph is its bipartite double cover. Our main tool is a geometric construction of the Steiner system $S(3,6,22)$.

fields
Combinatorics
pdf ↗ page arXiv:2611.07407 ↗

We compare transversal and pattern thresholds for two spanning structures. For every fixed $k\ge 3$ and $α>0$, we prove that every sufficiently large $n$-vertex $k$-graph collection with minimum codegree at least $(1/2+α)n$ contains a tight Hamilton cycle with any prescribed color pattern. Hence the transversal and pattern thresholds for tight Hamilton cycles coincide asymptotically. For spanning trees, we show that for every $α>0$ there exists $c=c(α)>0$ such that every $n$-vertex graph collection with minimum degree at least $(1/2+α)n$ contains every $n$-vertex tree with maximum degree at most $n^c$ and any prescribed color pattern. On the other hand, we construct spanning trees of maximum degree $\sqrt n+1$ whose pattern threshold exceeds their transversal threshold by a positive linear term.

comments
20 pages, 1 figure
fields
Combinatorics
pdf ↗ page arXiv:2611.07421 ↗

Let $ρ(G)$ denote the signless Laplacian spectral radius of a graph $G$, and let $\mathcal C_{n,m}$ be the family of connected graphs of order $n$ and size $m$. A graph is $Q$-maximal if it maximizes $ρ(G)$ over this family. Motivated by a question of Chang and Tam (2011), we study this extremal problem using scalar identities and coronal methods for threshold graphs. From the Perron equations, we derive the identity $S_G(λ)=T_{a,b}(λ)$ at $λ=ρ(G)$, where $a$ and $b$ are the total numbers of zero-vertices and one-vertices. This yields a one-point comparison principle: for a connected threshold graph $H$ with the same block totals, $S_H(z)>T_{a,b}(z)$ at some $z\ge n$ implies $ρ(H)>z$, allowing spectral comparison without estimating individual Perron coordinates. We prove that every $Q$-maximal threshold graph has character at most three, so its vertex-sequence contains at most three zero–one block pairs. We also show that the quasi-star graph is $Q$-maximal whenever $n\ge5$ and $n-1\le m<(n^2+10n-8)/8$. For certain dense parameter pairs, we construct connected threshold graphs with signless Laplacian spectral radii strictly larger than those of both the quasi-star and the connected quasi-complete graph. Thus the corresponding two-candidate assertion for connected graphs fails in general.

fields
Combinatorics
pdf ↗ page arXiv:2611.07429 ↗

A polytope $P$ is decomposable if it can be written as the Minkowski sum of two polytopes that are not homothetic to it. It is well known that decomposable $d$-dimensional polytopes with $2d$ or fewer vertices are the simplicial prisms. More recent work shows that, with two exceptions in low dimensions, decomposable $d$-dimensional polytopes with exactly $2d+1$ vertices are either pentasms or capped prisms. All these examples have a line segment for a summand. In this paper, we extend this conclusion to decomposable $d$-polytopes with exactly $2d+2$ or $2d+3$ vertices. In both cases, all such polytopes, with three exceptions, have a line segment for a summand. The exceptional cases are specified in detail.

fields
Combinatorics
pdf ↗ page arXiv:2611.07472 ↗

Let $Γ$ be a nontrivial group that acts on a graph $G=(V,E)$. The total distinguishing number $D''_Γ(V\cup E)$ (distinguishing index $D'_Γ(E)$) of the action is the least number of labels required for a total labeling (edge labeling) of $G$ so that the only elements of $Γ$ that preserve the labeling belong to the pointwise stabilizer of $V\cup E$ (pointwise stabilizer of $E$). We obtain upper bounds of these invariants for Cayley graphs and under various group actions, including free actions, wreath product actions, and coordinatewise actions on Cartesian powers of graphs. Our first main result shows that, if $Γ$ acts faithfully on $G$, with $|V|\geq 2, |E|\geq 1$ and the natural action of the wreath product $W=Γ\wr S_k$ on $V^k$ induces an action of $W$ on the Cartesian power $G^{k}=(V^{k},E_k)$, then $ D''_{W}(V^k\cup E_k)\le D''_Γ(V\cup E)+k-1 $ for any $k\geq 2$. We also establish the sharpness of several of our bounds through examples involving cyclic, symmetric, and alternating groups. Moreover, we determine these distinguishing invariants for the symmetric group action on Hamming graphs.

comments
13 pages, 4 figures. An extended abstract of a part of this work was presented at the Fourth Meru Combinatorics Conference 2026. Comments are welcome
fields
Combinatorics, Group Theory
pdf ↗ page arXiv:2611.07497 ↗

For a simple graph $G$, the partial Petrial polynomial $P_G^{\times}(z)$ enumerates all diagonal perturbations of its adjacency matrix over $GF(2)$ by rank. We derive a rooted attachment identity and apply it to the tree $R_m$ obtained by joining a new vertex to the middle vertex of each of $m$ disjoint copies of the five-vertex path. We prove that the nonzero coefficient sequence of $P_{R_m}^{\times}(z)$ is strictly unimodal for every $m\ge1$, but is log-concave if and only if $1\le m\le3$. Thus the $21$-vertex tree $R_4$ provides an explicit strictly unimodal, non-log-concave example. For $m\ge5$, the unique mode occurs at degree $\lfloor(149m+85)/32\rfloor$. More generally, for any fixed path rooted at an internal vertex, joining a new vertex to the roots of sufficiently many disjoint copies produces a tree whose partial Petrial polynomial is not log-concave.

comments
28 pages, 1 figure
fields
Combinatorics
pdf ↗ page arXiv:2611.07503 ↗

We prove that adjacency algebras of Jordan schemes over fields of characteristic zero are semisimple. We then classify the adjacency algebras of thin Jordan schemes over the fields of real numbers or complex numbers. The simple summands of the adjacency algebras over these fields are restricted to specific types of simple Jordan algebras.

comments
19 pages
fields
Combinatorics
pdf ↗ page arXiv:2611.07527 ↗

In \cite{R81}, Read obtained an explicit formula for the generalized chromatic polynomials of cycle graphs using the classical deletion-contraction principle. In this paper, we take a linear-algebraic approach and present a new proof of Read's formula by realizing the generalized chromatic polynomials of cycle graphs as traces of certain linear operators. As an application of this approach, we also derive explicit formulas for the generalized chromatic polynomials of uni-cyclic graphs.

fields
Combinatorics
pdf ↗ page arXiv:2611.07568 ↗

For a graph $G$ of order $N$, an integer $k$ with $0\leq k< N$ and a real number $p$ with $0\leq p\leq 1$, let $p(G,k)$ and $t(G,p)$ be the probabilities that $G$ is disconnected when $k$ vertices are removed randomly and when each vertex is removed independently with probability $p$, respectively. We deduce a formula of $p(G,k)$ for Harary graphs and show that the phase transition on $p(G,k)$ and $t(G,p)$ occurs in the class that has logarithmical degree. We also perform simulations on $p(G,k)$ for the Hypercube and randomly generated graphs of the same scale, which suggests that the regularity is a key structural feature that guarantees nice performance on $p(G,k)$. In particular, the Hypercubes have better performance than regular graphs, on average. Surprisingly, Harary graphs, known to be super-connected, have neither better performance than Hypercubes nor even than the average of regular graphs.

comments
17 pages,3 figures
fields
Combinatorics
pdf ↗ page arXiv:2611.07581 ↗

We obtain an explicit evaluation of the determinants that appear in the denominators of the integrands for the iterated integrals expressing special values of the spectral zeta function of the noncommutative harmonic oscillator at integers greater than 1. The determinants in question are those of cyclic tridiagonal matrices; their explicit evaluation is carried out, in addition to the inclusion-exclusion principle, conceptionally using combinatorial arguments based on monomer-dimer systems. This explicit evaluation serves as a core component in the study of arithmetic properties such as the modularity of the special values of the spectral zeta function.

comments
13 pages
fields
Combinatorics, Mathematical Physics, Classical Analysis and ODEs, Number Theory
pdf ↗ page arXiv:2611.07626 ↗

Adinkras are highly structured bipartite graphs that encapsulate 1D supersymmetry representations. We classify hypercube Adinkra heights with one, two, three or four edge colors using a graph-theoretic notion of equivalence. For the 990 distinct height functions in the case of four edge colors, we show that these classes match modified twisted character classes bijectively and provide a finer classification than spectral Height Yielding Matrix Numbers (HYMNs). We also relate our classification to the physics of the two-dimensional chiral and twisted chiral supermultiplets.

fields
Combinatorics
pdf ↗ page arXiv:2611.07637 ↗

We introduce a sliding procedure on diagrams which splits the standard Young tableaux of a corner-crossed diagram, in particular of a skew shape, along a finite binary tree whose leaves are anti-ribbon diagrams. Combined with standardization, this yields a decomposition of semi-standard Young tableaux into tableaux of anti-ribbon shape, indexed by standard tableaux and governed by their descent sets. As applications, we obtain expansions of Schur functions and Schur multiple zeta functions in terms of their anti-ribbon type analogues; at the level of Schur functions, this is a tableau-level form of Gessel's fundamental quasisymmetric expansion.

comments
24 pages
fields
Combinatorics, Number Theory
pdf ↗ page arXiv:2611.07715 ↗

A non-complete vertex-transitive graph is called 2-distance-primitive if a vertex stabilizer acts primitively on both the first and second distance neighborhoods of the vertex. Recently, Ruan, Liu and Jin [Discrete Math. 347 (2024), 113978] showed that every connected 2-distance-primitive graph of valency at most 9 either belongs to certain known families or satisfies a collection of further restrictions, with the socle of a vertex stabilizer isomorphic to $\mathrm{PSL}(2,7)$ or $\mathrm{PSL}(2,8)$. In this paper, we resolve these remaining cases and thereby obtain a complete classification of connected 2-distance-primitive graphs of valency at most 9. We also give a short proof of the main result of Ruan, Liu and Jin and identify additional graphs omitted from their list.

fields
Combinatorics
pdf ↗ page arXiv:2611.07733 ↗

We give a closed boundary formula for multiplication by power sums in the dual stable Grothendieck basis. The off-diagonal coefficients are signed binomial coefficients supported on corner-connected broken ribbons, which are skew diagrams connected through edges or vertices and containing no square of four boxes. For such a diagram, the smallest power-sum index with a nonzero coefficient is the number of boxes plus the number of edge-connected components minus one. All larger indices give nonzero coefficients when the boundary statistic is positive and zero coefficients when it vanishes. The diagonal coefficient is minus the number of removable corners of the indexing partition. The formula evaluates the stable specialization of Nguyen's Murnaghan–Nakayama rule for K-k-Schur functions. We give two independent proofs, one using Newton's identity and the dual Pieri rules and the other using determinantal formulas and Schur complements.

comments
21 pages
fields
Combinatorics
pdf ↗ page arXiv:2611.07795 ↗

For a hypergraph \(G\), let \(χ(G)\) denote its chromatic number and \(λ(G)\) the maximum, over pairs of distinct vertices, of the minimum size of an edge cut separating the pair. A hypergraph is \(3\)-critical if its chromatic number is three and every proper subhypergraph is two-colorable. In this paper, we give a recursive characterization of the \(3\)-critical hypergraphs with \(λ(G)\le2\). We prove that these are precisely the hypergraphs generated from ordinary odd cycles and an explicit family of labeled-tree hypergraphs by admissible Toft splits. This characterization includes a necessary and sufficient condition for a prescribed split to preserve \(λ\le2\). The condition involves only cuts in the first factor and the two traces on the split edge. More generally, we establish an exact cut formula for splits at degree-\(k\) vertices of \((k+1)\)-critical hypergraphs, for every \(k\ge2\). We further prove that the smallest member of this critical class outside the family generated by hyperwheels and both forms of the Hajós join has eight vertices and is unique up to isomorphism. Together, these results resolve two issues left open by Schweser, Stiebitz, and Toft in \emph{Journal of Combinatorics} (2022): the characterization of hypergraphs with \(λ(G)=2\) and \(χ(G)=3\), and whether hyperwheels and Hajós joins suffice to generate all \(3\)-critical hypergraphs in this connectivity range.

fields
Combinatorics
pdf ↗ page arXiv:2611.07856 ↗

Missing integers in partitions have been studied by many authors. A missing integer of a partition is a positive integer smaller than its largest part that does not occur as a part. We investigate parity biases for the number of missing integers in partitions and overpartitions. For every integer $c\ge2$, we prove that the coefficients $a_c(n)$ of $(-cq;q)_\infty/(-q;q)_\infty$ are eventually strictly positive and provide an effectively computable threshold. The proof uses the circle method, combining a saddle-point analysis on the major arc with explicit uniform estimates on the minor arcs. Specializing to $c=2$ and $c=3$, and verifying the remaining finite cases, we show that partitions and overpartitions of $n$ with an even number of missing integers outnumber those with an odd number for $n>34$ and $n>27$, respectively. We further prove that, for $n>10$, the difference between the numbers of ordinary partitions with an even and an odd number of missing integers is smaller than the number of partitions of $n$ into distinct parts. These results confirm three conjectures of Bhoria, Eyyunni and Santra. We also give a combinatorial interpretation of $a_c(n)$ for every integer $c\ge4$ in terms of colored partitions, generalizing the partition and overpartition interpretations of Bhoria, Eyyunni and Santra for $c=2$ and $c=3$, respectively.

comments
34 pages
fields
Combinatorics, Number Theory
pdf ↗ page arXiv:2611.07908 ↗

For an \(n\)-vertex graph \(G\) and an integer \(m\ge n\), let \(f'(G,m)\) be the minimum number of distinct color sets used by proper colorings from \(m\)-element lists that are disjoint at distinct nonadjacent vertices. A clique-partition assignment gives a common list to the vertices of each part of a clique partition and uses disjoint lists on different parts. In this paper, we construct a graph on \(20\) vertices for which these assignments do not attain \(f'(G,20)\), and a fixed graph on \(25\) vertices for which they fail to attain the minimum for every allowed list size. For every triangle-free graph \(G\) on \(n\ge2\) vertices, we prove \[ f'(G,m)=\binom{m}{2}^{ν(G)}m^{n-2ν(G)} \qquad (m\ge n), \] where \(ν(G)\) is the size of a maximum matching, and characterize all minimizing assignments in terms of maximum matchings. For clique blow-ups of each fixed odd cycle of length at least five, we determine the minimum up to constant factors uniformly over the blow-up size and all allowed list sizes, and obtain an exponential separation from the clique-partition value. We also prove that a noncomplete graph \(G\) satisfies \(f'(G,n)=n^2\) if and only if all edges of its complement have a common endpoint. These results disprove the general clique-partition conjecture of Gravier, Maffray, and Mohar (\emph{Discrete Mathematics}, 2003) and its extension by Wu and Zhang (\emph{Applied Mathematics Letters}, 2006), while establishing both triangle-free conjectures.

fields
Combinatorics
pdf ↗ page arXiv:2611.07924 ↗

We study $\mathbb F_q$-linear sets whose positive point weights belong to a prescribed finite set $A$. The integer $e=\gcd(A)$ always divides the rank of the defining subspace, but the divisibility does not by itself imply linearity over $\mathbb F_{q^e}$. We show that it does in a broad parameter range. Let $L_U$ be an $\mathbb F_q$-linear set of rank $N$ spanning $PG(h-1,q^m)$, $h\ge2$, without being equal to it, and assume that $q\ge m$ and that the minimum weight of its points is $a\ge2$. Combining the theorem of Csajbók, Marino and Pepe on the maximum field of linearity with the lower bound of Adriaensen and Santonastaso for the size of a linear set containing a point of weight one, we prove that $L_U$ is also defined by an $\mathbb F_{q^t}$-subspace, for some divisor $t$ of $m$, whose $\mathbb F_q$-dimension is exactly $N+t-a$. Under a natural gap condition excluding divisors of $m$ in the interval $(e,N-e]$, this forces $a=e$ and the defining subspace to be $\mathbb F_{q^e}$-linear. Through geometric duality, the point weights of the dual of a system are the deficiencies $m-w$ of the rank weights $w$ of the associated code. For rank-metric codes in which the subcode generated by the codewords of nonmaximum rank contains codewords of maximum rank, we show that deleting suitable coordinates yields the $\mathbb F_q$-expansion of a code over $\mathbb F_{q^t}$. We show that the codes in which the codewords of nonmaximum rank generate a subcode without codewords of maximum rank are exactly the direct sums of a simplex code and a code of length smaller than $m$. We derive field-reduction results for two-weight and three-weight rank-metric codes, and show that no gap condition is needed when $m$ divides the length or when the code has only two nonzero weights. Finally, we study two-dimensional codes of length at most $m$ and provide families of three-weight examples.

fields
Combinatorics
pdf ↗ page arXiv:2611.08044 ↗

We prove that there is a unique tight $4$-design in the real Grassmannian $G_{2,4}$, up to the action of $O(4)$. Using symmetric association schemes, we also construct an $N$-point $4$-design in $G_{2,4}$ for every even integer $N\geq10$. For symmetric association schemes of order less than $21$, we classify the pairs of rank-three primitive idempotents defining injective maps into $G_{2,4}$; their images are Grassmannian $4$-designs precisely for four schemes of order $16$.

comments
24 pages
fields
Combinatorics, Metric Geometry
pdf ↗ page arXiv:2611.08047 ↗

The Hirsch problem asks how large the diameter of the dual graph of a simplicial complex can be, in terms of its number of vertices. For normal complexes, the classical bounds are the Larman-Barnette bound, which is linear for fixed dimension, the Kalai-Kleitman bound, which is quasi-polynomial, and the Adiprasito-Benedetti bound n-d, which however holds only for flag complexes. After sharpening the Larman-Barnette and the Kalai-Kleitman bounds, we study how the three bounds deteriorate when flagness or normality is weakened. We give three ways to quantify the distance from flagness, and in each case we bound the diameter in a way that interpolates between the Adiprasito-Benedetti and the Larman-Barnette bounds. We also quantify the distance from normality by a parameter q, and obtain a bound that is a polynomial of degree q in n. Finally, we conjecture a statement on flag layered families that would imply an interpolation between the Adiprasito-Benedetti and the Kalai-Kleitman bounds.

comments
22 pages
fields
Combinatorics, Metric Geometry
pdf ↗ page arXiv:2611.08071 ↗

In their 1998 precursor to their celebrated paper ``Recounting the rationals,'' Calkin and Wilf listed nine open problems related to the function $b_{k,\ell}(n)$, the number of partitions of $n$ into powers of $\ell$ with each part appearing at most $k$ times. In this short paper, we answer five of their eight remaining open problems. As a consequence of our results, we extend and generalise results of both Carlitz and Reznick, and answer an open question of Reznick.

comments
8 pages
fields
Combinatorics, Number Theory
pdf ↗ page arXiv:2611.08109 ↗

Monadically dependent graph classes are those in which first-order logic cannot encode all graphs. They include all nowhere dense classes and all classes of bounded twin-width. We show that first-order model checking is fixed-parameter tractable on every monadically dependent graph class. Together with the known hardness result for hereditary classes that are not monadically dependent, this gives a complete characterization of the hereditary graph classes with fixed-parameter tractable first-order model checking. This resolves a central open problem in algorithmic model theory. Our proof combines ideas from structural graph theory (graph decompositions), finite model theory (locality of first-order logic), and model theory (counting types). In particular, we show that in monadically dependent classes, tuples of fixed length have only $|U|^{O(1)}$ types of fixed quantifier rank over any set of parameters $U$.

fields
Combinatorics, Logic
pdf ↗ page arXiv:2611.08143 ↗

We generalize the concept of token graphs, and we call them supertoken graphs. In this work, we consider the variant in which tokens are distinguishable (for instance, by color) and at most one token is allowed on each vertex. We formally define this kind of supertoken graphs and present their basic properties. We study their diameter, in which cases they are Cayley graphs, and their spectra. Moreover, we provide bounds or exact values for the independence, the vertex cover, the clique, and the chromatic numbers of these graphs. Finally, we characterize the cases in which the supertoken graphs of complete graphs are planar.

fields
Combinatorics
pdf ↗ page arXiv:2611.08186 ↗

We investigate the optimal asymptotic rate $R_q$ of perfect $q$-hash codes, a fundamental extremal problem in combinatorics, information theory, and computer science. Let $B_q=-\log_2(1-q!/q^q)/(q-1)$ denote the classical probabilistic lower bound. We prove that $R_q=B_q$ if and only if $q=2$, establishing an explicit strict improvement for every $q\ge3$ and settling the first assertion of a conjecture of Xing and Yuan (Combinatorica 43 (2023), 347–366). More significantly, we show that, for every $η>0$, $R_q\ge B_q+\exp(-(2+η)q)$ for all sufficiently large integers $q$. This improves the previously known increments of scale $\exp(-3q+O(\ln q))$ by an exponential factor. We also establish a structural limitation of the concatenation method. For pairwise uniform inner codes of subexponential length, the resulting rate bound exceeds $B_q$ by at most $\exp(-3q+o(q))$. Thus, exponentially long inner codes are necessary to achieve the new improvement scale within this framework. Our proofs combine Fourier analysis, high-dimensional simplex codes, additive permutation counts, and entropy bounds for hypergraph matchings. We further strengthen lower bounds from short inner codes and improve several known numerical estimates.

fields
Combinatorics, Information Theory
pdf ↗ page arXiv:2611.08230 ↗

Akbari, Cameron, and Khosrovshahi proposed two questions concerning the null space and the image space of graphs. Although these problems have been studied in several cases, they have not yet been fully resolved. In this work, we present constructions of graphs that satisfy both the problems.

fields
Combinatorics, Rings and Algebras
pdf ↗ page arXiv:2611.08244 ↗

We compute Cunningham's canonical decomposition of a symmetric submodular function on $n$ elements in $O(n^3)$ time. The decomposition represents all minimum cuts in linear space. For the pair $s,t$ produced by Queyranne's algorithm, we construct in $O(n^2)$ time a digraph representing exactly the minimum $st$-cuts of the function. Using this construction in Chekuri and Xu's decomposition algorithm gives the cubic bound.

fields
Combinatorics, Optimization and Control
pdf ↗ page arXiv:2611.08253 ↗

In this paper, we establish a matrix criterion for a space to be homeomorphic to a closed Euclidean ball. As applications, we prove that the normalized space of $K$-Lorentzian polynomials on an arbitrary open convex cone $K$ is homeomorphic to a closed Euclidean ball, giving an affirmative answer to Brändén and Leake's question, and we also settle Brändén's conjecture that the normalized space of homogeneous multiaffine real stable polynomials is homeomorphic to a closed Euclidean ball. In addition, our results show that the normalized space of homogeneous $K$-stable polynomials of degree $d$ is homeomorphic to a closed Euclidean ball.

comments
31 pages
fields
Combinatorics
pdf ↗ page arXiv:2611.08265 ↗

Eldan's level-$1{:}3$ inequality is a central tool in his second-order correlation inequalities for increasing Boolean functions. His proof is going to Gaussian space and uses an entropy estimate for second Hermite tensors. We give a direct proof on the discrete hypercube. For increasing real-valued functions $f,g:\{0,1\}^n\to\mathbb R$, we prove $$ W_3[f,g]\ge -2W_1[f,g]\log\!\left(\frac{\sqrt{\mathrm{Var}(f)\mathrm{Var}(g)}}{W_1[f,g]}\right). $$ The proof combines complex hypercontractivity with the discrete Riesz transforms. We also give a short proof of a level-$1:d$ inequality obtained by Keller and Kindler. As an application, we recover Eldan's correlation inequality with a slight improvement. If $f,g:\{0,1\}^n\to\{0,1\}$ are increasing and $g$ is antipodal, then $$ \mathrm{Cov}(f,g)\ge \frac1{24}\,\frac{\sum_i\mathrm{Inf}_i[f]\mathrm{Inf}_i[g]}{\sqrt{\log\!\left(4e\sqrt{\mathrm{Var}(f)\mathrm{Var}(g)}/(\sum_i\mathrm{Inf}_i[f]\mathrm{Inf}_i[g])\right)}}, $$ where the square-root logarithm cannot be improved as shown in the OR–majority example.

comments
21 pages, comments welcome!
fields
Combinatorics, Classical Analysis and ODEs, Functional Analysis, Probability
pdf ↗ page arXiv:2611.08289 ↗

Inspired by the work of Poljak, Bašić, Damnjanović, Stevanović, and Stošić used the Kissat SAT solver in a recent paper to systematically determine the small cyclic and ordered Ramsey numbers of several families of graphs. By studying the patterns in their data, Bašić et al. formulated 15 conjectures on cyclic and ordered Ramsey numbers. In this paper, we confirm 9 of their conjectures, trace confirmations of 2 more to existing literature, and, perhaps surprisingly, disprove another one.

comments
21 pages, 3 figures, 2 tables
fields
Combinatorics
pdf ↗ page arXiv:2611.08375 ↗

We prove a Dirac-type condition for an $r$-partite graph to admit a fractional clique decomposition: every sufficiently large $r$-partite graph with $n$ vertices in each part and partite minimum degree at least $(1 - \frac{1}{30r^2}) n$ admits a fractional $K_r$-decomposition, subject to necessary divisibility conditions. This implies the same upper bound on the asymptotic minimum degree threshold for $r$-partite graphs to admit a $K_r$-decomposition, using a well-known reduction.

comments
33 pages, 2 figures
fields
Combinatorics
pdf ↗ page arXiv:2611.08432 ↗

We determine the sharp order of discrepancy in Tusnády's problem in every fixed dimension $d\ge 2$. The largest combinatorial discrepancy of $N$ points in $\mathbb{R}^d$ with respect to axis-parallel boxes is $Θ_d((\log N)^{d-1/2})$, extending Wei's planar result and matching Nikolov's upper bound. More strongly, with high probability, this lower bound holds simultaneously for every subset obtained by deleting at most a sufficiently small fixed fraction of a uniform $N$-point sample. We apply this geometric estimate to online thinning of two independent samples of size $n$ from an unknown common distribution, where observations must be kept or discarded upon arrival. For every fixed dimension $k\ge 1$ and expected deletion budget $ρn$, with $0<ρ<2$, we prove that the minimax expected prefix discrepancy over lower orthants is $Θ_{k,ρ}((\log n)^{k+1/2})$. In dimension one, the optimal terminal discrepancy is instead $Θ_ρ((\log\log n)^{3/2})$, giving a sharp separation between the prefix and terminal objectives. For $k\ge 2$, we obtain terminal bounds differing by a factor of $\log n$. The lower bounds are based on Gaussian small-ball estimates for box-incidence matrices, while the upper bounds use stationary Metropolis thinning and, in one dimension, a terminal cancellation procedure.

comments
40 pages, 5 figures
fields
Combinatorics, Probability
pdf ↗ page arXiv:2611.08455 ↗

Analysis of PDEs48

Registration maps propagate labels, measure local volume change, and map results back through their inverses; requiring invertibility. Voxelwise checks can miss folds between grid points. We study the canonical continuous extension of a nodal displacement field by trilinear interpolation and give a test that proves its Jacobian determinant positive in every grid cell or rejects the map. The test is rigorous under IEEE 754 binary64 arithmetic (correct rounding, no contraction). With a fixed boundary, certification implies a bi-Lipschitz homeomorphism of the image domain. We introduce SeqVP, built around this certificate. A proximal Jacobian-determinant-curl formulation yields inner problems whose penalties on changes in the map, Jacobian determinant, and curl vanish to first order at the current iterate. The method decreases a fixed outer energy up to a floating-point resolution guard and certifies every committed iterate. After certified initialization, stopping at any time returns a folding-free map. A Poisson-preconditioned gradient method with Jacobian-aware step prediction and affine-scaling damping, initialized by certified handoff from an uncertified binary32 pre-optimizer, handles a 1 mm brain MRI with about 2e7 unknowns in 90 s on one GPU. On 49 held-out OASIS pairs, SeqVP achieves mean Dice 0.821 over 35 structures and 0.889 over 4 tissue classes, with certified zero folding. Gains are 0.022 and 0.025 over ANTs SyN, and 0.022 and 0.023 over uniGradICON with instance optimization at equal time (paired, Holm-adjusted p < 0.001), but maps are rougher. SeqVP is level with ANTs on LPBA40, behind it on IBSR18, and ahead of all label-free baselines on 20 large-deformation pairs. On fields from nine methods, the certificate finds non-positive-Jacobian cells in every ANTs SyN map (median 10, missed by voxelwise checks) and in maps from NiftyReg, ConvexAdam, uniGradICON, and TransMorph.

comments
51pages
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.06954 ↗

This Ph.D. thesis concerns the blow-up analysis, and more precisely, the asymptotic decomposition into profiles of solutions to nonlinear partial differential equations. In Chapter 1, we present a unifying framework for finding profile decompositions in Hilbert spaces using the notion of dislocation spaces. The next chapters are then devoted to my publications. In the second chapter, we construct solutions to the focusing, energy-critical, nonlinear wave equation in dimension 4 and 5, exhibiting finite-time Type II blow-up precisely at the origin with a single profile and a prescribed polynomial blow-up rate of order $-1-ν$, where $ν> 1$ in 4D and $ν> 3$ in 5D. The major difference between dimension 4 and 5 consists in the renormalization procedure. In 4D, we essentially follow the Krieger-Schlag-Tataru scheme developed for the 3-dimensional equation, where the polynomial structure of the nonlinearity permits the use of simple algebraic manipulations to control error terms. By contrast, the 5D case requires a modified setup due to the lower regularity of the nonlinearity, which complicates the treatment of nonlinear error terms. In Chapter 3, we provide a modern presentation of classical Lp and Schauder estimates for parabolic equations. Then, in Chapter 4, we consider finite-time and $k$-equivariant solutions to the harmonic map heat flow from a disk to a sphere. Under general time-dependent, $k$-equivariant and smooth boundary data, we prove that the profile decomposition at the singular time contains only one bubble. The method relies on the maximum and comparison principle. We also prove a global existence criterion and exhibit solutions blowing up in infinite time for any degree of equivariance. Finally, the appendix presents the profile decomposition for the harmonic map heat flow proved by Struwe and Qing using energy arguments and the Palais-Smale condition.

comments
This thesis contains more details than my publications, hence it is useful to have it on arXiv
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.06959 ↗

We study the interior regularity of convex solutions to $$\det D^2u = f(u)$$ in planar domains, with no regularity or monotonicity assumptions on $f$. In particular, when $f$ is bounded between positive constants, we show that $u$ is locally $C^{1,\,1}$, and that its level sets near a critical point are asymptotically ellipses. Our analysis relies on a new monotone quantity for a second-order flow satisfied by the support functions of the level sets of $u$. As a byproduct of our methods, we obtain Liouville theorems that generalize a classical result of Jörgens, and include a complete classification of global solutions in 2D to an obstacle problem studied by Jin-Tu-Xiong. As another application, we obtain regularity and rigidity results for rotating-wave solutions of the semigeostrophic equations from meteorology.

comments
43 pages, 2 figures. Comments are welcome!
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.06978 ↗

We prove an Obukhov-Corrsin-Yaglom-type law for passive scalars advected by a family of random $C^α$ velocity field proposed in \cite{JDEMS} as a kinematic model of turbulent fluids. Up to a technical assumption on one parameter, the expectation of the squared $L^2_t H^γ_x$ norm of solutions to the passive scalar equation is bounded by the squared $L^2$ norm of the deterministic initial datum, uniformly in the viscosity, whenever $α+ 2γ< 1$. The proof is based on a new insight in the theory developed by \cite{FM} to prove Richardson's dispersion and anomalous dissipation for the same model.

fields
Analysis of PDEs, Probability
pdf ↗ page arXiv:2611.07076 ↗

We construct a divergence-free velocity field $\mathbf{u}(\mathbf{x},t)$ on $\mathbb{T}^{n}$ that is a genuine function of position and carries both defining properties of turbulent transport: Richardson pair dispersion and a dissipative anomaly. No synthetic field has carried both, because the frame that gives the first destroys the field the second requires. The key is a dichotomy: deforming a spectrally localised pattern smears it across scales, whereas pushing its generating centres forward and re-convolving leaves the spectral envelope exactly invariant --- so transporting the eddy centres preserves both cascade and frame. In the field the finite-size Lyapunov exponent obeys $ζ(R)=h(R)-1$ threshold by threshold within the scaling range, with a sweeping deficit $Δ_{\rm sw}=+0.038\pm0.019$, consistent with zero, against $+0.091\pm0.019$ for the frozen field on the same data, a gap that doubles at Navier–Stokes intermittency. Companion papers prove that this family shows Richardson dispersion and a dissipative anomaly (Fefferman and Mayboroda 2026), and that a scalar advected by this flow obeys Yaglom's relation (De Lellis 2026), at small Kubo number. Trajectories forget their starting point as the smallest eddy is removed: the median separation of pairs released from one point with independent Brownian kicks varies as $κ_{p}^{0.0333\pm0.0038}$ over 22 decades of particle diffusivity, against $κ_{p}^{1/2}$ for a smooth control. A scalar in the same field closes the exact scale-by-scale variance budget, four independently measured terms balancing the dissipation to 15\% over 1.2 decades with no adjustable constant, and reaches its inertial–convective asymptote $S_{L}=-χr$ to $0.9995\pm0.0018$ once the measured diffusive transfer is restored, while under an imposed mean gradient $χ$ changes by 14\% over a factor 5.8 in $κ$.

comments
The code, the analysis scripts that produce every figure and numbers in the paper, the run configurations, the data behind the figures and the supplementary movies are available at https://github.com/R1967J/kinematic-field. A single self-contained generator of the velocity field kinematic_field.py
fields
Analysis of PDEs, Numerical Analysis
pdf ↗ page arXiv:2611.07121 ↗

Let $ C_ω$ be the cone generated by a proper $C^2$ domain $ω\subset\mathbb S^{N-1}$. We study positive homogeneous solutions of the fractional $p$-Laplace equation in $C_ω$ with zero exterior values. For $p\ne2$, the operator is nonlinear and not uniformly elliptic. For every $p>1$, we identify two extremal homogeneities $β^+>0>β^-$, prove that every fundamental solution, if one exists, has one of these exponents, and construct homogeneous eigenfunctions whose eigenvalues have signs determined by the position of the homogeneity relative to $β^\pm$. As an application, we obtain the Lane-Emden threshold $q^*=p-1+sp/β^+$ for every $p>1$: nontrivial nonnegative supersolutions of $(-Δ_p)^s u\ge u^q$ in $C_ω$ do not exist for $0<q<q^*$, whereas homogeneous supersolutions exist for $q>q^*$. We also establish Hopf-type lower bounds at conical boundary points with the optimal exponent. When $p-1>sp$ , we construct both fundamental solutions, one singular at the vertex and the other vanishing there. For $p=2$, these solutions are known; our method recovers them for every $s\in(0,1)$. The homogeneous eigenfunctions also serve as barriers, yielding boundary decay and global Hölder regularity for $(-Δ_p)^s u=f$ in bounded Lipschitz domains, with zero exterior values and $f$ continuous up to the boundary. The admissible Hölder exponents are determined by the exterior cones. All results hold for every $N\ge2$; in particular, $N>sp$ is not needed.

fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07157 ↗

We study the inhomogeneous logarithmic Schrödinger equation \[ i\partial_tu+Δu+|x|^{-1}u\log|u|^2=0, \qquad (t,x)\in\R\times\R^N,\quad N\ge2. \] For every initial datum in $H^1(\R^N)$, we establish global existence of weak solutions. We also construct explicit positive stationary solutions and solve the associated Nehari minimization problem. For $ω\ge0$, the minimizer is unique up to {a constant} phase and is given by \[ e^{\frac{N-1}{2}(1+\sqrt{1+ω})} e^{-(1+\sqrt{1+ω})|x|}, \] whereas for $ω<0$ the infimum is zero and is not attained. Finally, we prove that the standing waves associated with these profiles are orbitally stable for $ω>0$ by a variational argument, and orbitally unstable for $-1<ω<-1+(N-1)^{-2}$ by combining spectral analysis, localized energy and mass estimates, and phase modulation.

comments
29 pages
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07167 ↗

We study energy methods for the time-fractional equation $\partial_t^α(u-u_0)-(a(t,x)u_x)_x+(F(t,x,u))_x=Q(t,x,u)$ on the real line, where $0<α<1$, the coefficient $a$ is measurable and bounded from above and from below by positive constants, $σ\partial_xF\ge0$, $σQ\le0$, and the data are in $L^1\cap L^\infty$. Using Yosida approximations of the kernel, an iteration on the $L^{2^m}$ norms gives the parabolic decay $\|u(t)\|_{L^\infty}\le C\|u_0\|_{L^1}t^{-α/2}$. In $L^2$ it gives the rate $t^{-α/4}$, which, already for the linear equation, improves the rate $t^{-α/5}$ obtained before by the classical energy method. For the convection–diffusion equation $\partial_t^α(u-u_0)-u_{xx}+(|u|^{q-1}u)_x=0$, $q>1$, the parabolic rate is optimal for $q\ge2$, and for $1<q<2$ the convection improves it. An $L^1$ time-regularity estimate of Bénilan–Crandall type gives the pointwise flux bound $|u|^{q-1}u-u_x=O(\|u_0\|_{L^1}t^{-α})$. This bound takes the place of the Oleĭnik estimate and yields the optimal rate $\|u(t)\|_{L^\infty}\lesssim t^{-α/q}$ for $1<q<2$. For nonnegative data, matching lower bounds hold for the $L^p$ norms.

comments
34 pages
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07185 ↗

In this paper, we study sign-changing normalized solutions of the Kirchhoff problem \[ \begin{cases} -\left(a+b\displaystyle\int_Ω|\nabla u|^2\,dx\right)Δu =|u|^{p-2}u+μu &\text{in }Ω,\\ u=0 &\text{on }\partialΩ, \end{cases} \] subject to the constraint \[ \int_Ω|u|^2\,dx=c, \] where $a,b>0$, $Ω\subset\mathbb R^N$ is a smooth bounded connected domain, $1\le N\le3$, $c>0$ is prescribed, and $μ\in\mathbb R$ is a Lagrange multiplier. We assume $2<p<\infty$ for $N=1,2$ and $2<p\le6$ for $N=3$. For every integer $k\ge1$, we obtain at least $k$ distinct pairs of sign-changing normalized solutions for sufficiently small $c>0$. We determine the small mass asymptotics of their energies, Lagrange multipliers, and rescaled profiles in terms of the distinct Dirichlet eigenvalues and their eigenspaces. When $2<p<2+\frac{8}{N}$, we obtain infinitely many such pairs for every prescribed mass $c>0$, with energies tending to $+\infty$. At $p=2+\frac{8}{N}$, this conclusion holds for every sufficiently small fixed mass. We also establish the existence of a least energy sign-changing normalized solution on general bounded domains for every prescribed mass when $2<p<2+\frac{8}{N}$. For sufficiently small mass, such a solution exists on general bounded domains when $p=2+\frac{8}{N}$ and on star-shaped domains when $p>2+\frac{8}{N}$. For any family of such least energy solutions, the rescaled profiles approach the $L^2$-unit sphere of the second Dirichlet eigenspace in $H_0^1(Ω)$ as $c\to0^+$. The finite multiplicity and least energy results include the Sobolev critical case $N=3$, $p=6$. The proofs combine a localized symmetric minimax construction, energy comparisons with one-signed critical points, and compactness arguments that preserve the sign-changing property.

fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07199 ↗

We consider a closed planar curve evolving by the free or the length-preserving elastic flow. We prove that the scale-invariant bound $L\int k^2\,ds\le 4π^2+8/3$ at the initial time prevents the entire curve from crossing any line between two times. For the free flow, the proof combines a monotonicity estimate for the normalized bending energy with a barrier that compares centroid motion to an expanding energy radius. For the length-preserving flow, a quantitative gradient estimate controls the total centroid displacement.

comments
12 pages
fields
Analysis of PDEs, Differential Geometry
pdf ↗ page arXiv:2611.07242 ↗

We establish the global pathwise well-posedness and a Freidlin–Wentzell large deviation principle for the three-dimensional viscous primitive equations driven by multiplicative Lévy noise represented by a compensated Poisson random measure. Under local Lipschitz and moment growth assumptions on the noise coefficient, we construct local martingale solutions and obtain a unique maximal pathwise solution with càdlàg trajectories in the $H^1$ energy space. Additional sixth-moment growth in $H^1$ and $L^6$ growth bounds on the baroclinic component of the noise yield global pathwise well-posedness. The proof combines the jump Itô formula with baroclinic and anisotropic energy estimates to rule out finite-time blow-up. Under further exponential integrability and structural assumptions on the noise coefficient, we establish a Freidlin–Wentzell large deviation principle in a strong path topology by the weak convergence approach.

comments
55 pages
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07274 ↗

We study global well-posedness and stability for Boltzmann equations with linear velocity deformation, allowing for arbitrarily large deformation strength. For diagonal dilatation with $-2<γ\leq0$ and cylindrical dilatation with $-3/2<γ\leq0$, with or without shear, we construct global homogeneous solutions for small data and prove their stability under small spatially inhomogeneous perturbations. In scaled variables, the perturbations remain uniformly controlled in deformation-adapted spaces and decay in mixed spatial–velocity norms, algebraically in the diagonal regime and exponentially in the cylindrical regime. For homogeneous simple shear with $-2\leqγ\leq0$, we prove global existence and uniqueness for nonnegative data with finite mass and Fisher information, without smallness or a finite-energy assumption. These solutions conserve mass and propagate additional Sobolev regularity. For spatially inhomogeneous simple shear with $-2<γ\leq0$, we establish global well-posedness for small data. The analysis combines commuting vector fields, dispersive and bilinear collision estimates, and integrability of the rescaled collision coefficients. For homogeneous shear, a positivity-preserving splitting transfers Fisher-information control to rough data, while a Lorentz-space estimate includes the endpoint $γ=-2$ in the global continuation argument.

fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07340 ↗

Let $T$ be a nonhomogeneous rooted tree equipped with a positive flow measure $m$. For $1<p<\infty$, consider the variable-coefficient Sobolev space on the nonhomogeneous flow-measure tree~$T$ \[ W_a^{1,p}(T,m) = \left\{ f\in L^p(T,m): \sum_{x\ne ρ}a(x)m(x)\lvert f(x)-f(π(x))\rvert^p<\infty \right\}, \] where $a$ is a positive variable coefficient, $ρ$ denotes the root and $π(x)$ denotes the parent of $x$. Using an averaging method over descendant levels, we prove the density of finitely supported functions in $W_a^{1,p}(T,m)$ for coefficients satisfying a weighted moment condition. As an application, the relationship between solutions on finite domains and global solutions is clarified for $p=2$.

fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07350 ↗

In this paper, we consider the following two-coupled nonlinear Schrödinger equations: \begin{equation*} \begin{cases} &-Δu+P(x)u=μu^3+βuv^2,\quad\text{in }\R^N,\\ &-Δv+Q(x)v=νv^3+βu^2v,\quad\text{in }\R^N,\\ &u,\,v\in H^1(\R^N) \end{cases} \end{equation*} where $N=2,3$, $P(x)$ and $Q(x)$ are potentials, $μ$ and $ν$ are positive coefficients and $β\in\R\setminus\{0\}$ is the coupling parameter. By developing the intermediate reduction arguments in [del Pino-Wei-Yao, Calc. Var., 2015], we prove the existence of infinitely many nonradial positive solutions to the above system under suitable regularity and decay conditions on $P(x)$ and $Q(x)$, {\bf but without any symmetry assumptions on them}. The scalar version of our result also gives {\bf a positive answer} to the first part of [Question~3, del Pino-Wei-Yao, Calc. Var., 2015] and our method might be helpful in studying other problems.

comments
Any comments are welcome!
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07359 ↗

Prékopa's theorem shows that, for jointly log-concave initial data, free transport preserves the log-concavity of spatial densities. This paper asks whether this property persists in the repulsive Vlasov–Poisson system. In dimensions two and higher, we construct smooth, strongly log-concave initial distributions whose spatial densities lose log-concavity at arbitrarily small positive times. For one-dimensional cold-ion data, we prove that the spatial density remains log-concave for all time.

fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07380 ↗

In this paper, we study an integral equation of Landau-Lifschitz-type involving the Riesz potential $$ u(x)=\ell+C_*\int_{\mathbb{R}^n}\frac{u(y)(|\nabla u(y)|^{2}+|u_k(y)|^2)-u_k(y)e_k}{|x-y|^{n-α}}dy, $$ where $\ell \in \mathbb{R}^k$ is a constant vector, $C_* \neq 0$ is a constant, $u\colon\mathbb{R}^n \to \mathbb{S}^{k-1}:=\{u=(u_1,u_2,\cdots,u_k) \in {\mathbb{R}^{k}};|u|=1\}$, $k \geq 1$, $n \geq 2$, and $0<α<n$. We provide a decay rate at which $|u_k|$ converges to zero under some mild conditions. This argument still works for the Ginzburg-Landau-type integral equation.

fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07391 ↗

We present a computational framework that combines machine-learning-based discovery with rigorous computer-assisted numerical certification and formal verification for strongly nonlinear partial differential equations. We demonstrate the approach for periodic gravity-capillary traveling water waves. Physics-informed neural networks are used as a search mechanism, with the wave profile, wave speed and, in continuation experiments, surface tension learned directly from the governing equations without training data. Through a staged training and refinement procedure, the neural networks achieve near-machine-precision residuals of the governing equations, even in strongly nonlinear regimes. A single formulation recovers high-amplitude Stokes waves, Crapper waves and Wilton ripples. Continuation from a high-amplitude Wilton ripple in surface tension reveals new high-amplitude candidate solutions. We freeze one candidate as an explicit finite representation and rigorously certify a uniform bound on its governing-equation residual using interval arithmetic, accounting for floating-point roundoff, Fourier truncation, aliasing and grid-to-continuum errors. This independently validates the neural-network accuracy. We formalize the certification argument in Lean, which machine-checks the deductions from certified numerical bounds to the residual theorem. Both the interval certification and formalization were developed with substantial assistance from a large language model, while the validity of the result remains independent of the language model: the numerical bounds are rigorously enclosed by interval arithmetic and the subsequent mathematical deductions are checked by the Lean kernel. Together, these steps provide a pipeline from neural discovery to rigorous numerical certification and machine-checked verification with auditable mathematical guarantees.

comments
29 pages, 6 figures
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07395 ↗

For microorganisms with run-and-tumble motility, kinetic models provide a well-established description of transport in smoothly varying environments. Sharp signal changes, however, can induce asymmetric interface responses. A motivating example is the response to dark-light boundaries: cells moving into the light largely maintain their direction, whereas those entering the dark are effectively scattered back. Through an interface-layer analysis of pathway-based kinetic equations, we formally derive effective interface conditions for the limiting run-and-tumble model. These velocity-nonlocal conditions capture non-adiabatic scattering near signal discontinuities, linking intracellular sensory dynamics to transport across sharp environmental boundaries. For a two-velocity model, they reduce to an explicit transmission matrix. We assess the interface law numerically using biologically motivated parameters for E. coli chemotaxis.

fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07409 ↗

Let $k\ge2$ be an integer. We prove that every minimizer of the Bethuel–Brezis–Coron relaxed Dirichlet energy in the signed meridional $k$-axial class, with continuous $H^1$ degree-zero data on the boundary of a ball, is smooth outside a finite subset of the symmetry axis. This extends the weight-one finiteness theorem of Hardt–Lin–Poon and resolves the finite-singularity question posed by Martinazzi–Rivière as Open Problem~A, in the fixed-boundary Bethuel–Brezis–Coron setting with continuous degree-zero data. The interior argument uses a distributional graph-boundary identity, height-preserving half-turn replacements, and exact square completions retaining the current cost. Cap-valued crosscuts and a finite incidence count exclude singular accumulation at the boundary poles.

comments
22 pages, 4 figures
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07414 ↗

We study the dispersive properties of the massive Dirac equation perturbed by an Aharonov-Bohm magnetic field, proving in particular dispersive and Strichartz estimates for the associated flow. This paper is the natural continuation of our previous work in the massless case [3], and the overall structure of the analysis remains similar. The presence of a nonzero mass term, however, breaks the scaling invariance of the equation and introduces new difficulties at several stages of the argument. In particular, the appearance of the spectral thresholds +m and -m leads to a genuinely different low energy regime, which requires a separate analysis of the singular angular component.

comments
44 Pages
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07445 ↗

This paper establishes interior $H^{s,p}$ regularity for weak solutions to fractional nonlocal equations in divergence form with general kernels. Under the $(δ, R)$-vanishing condition on the coefficients, we prove that for $p>2$, if the source term $f\in L^p_{loc}$, then the solution $u\in H^{s,p}_{loc}$. In contrast to Byun and Kim \cite{BK2025}, our result holds in the borderline case for $f$ without requiring differentiability assumptions. Our approach simplifies previous methods by using the stability properties of the fractional gradient to derive the good-inequality directly, thereby avoiding reliance on maximal functions. The proof integrates Vitali's covering lemma, Hardy-Littlewood maximal theory, self-improving estimates, and compactness arguments. This work extends classical $W^{1,p}$ regularity to fractional nonlocal settings under minimal kernel assumptions.

fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07606 ↗

Let $n\geq2$, $0<β<n$, and let $V$ be a continuous nonnegative potential on $\R^n$ with a positive lower bound outside a ball. Under the constraint $\int_{\R^n}(|\nabla u|^n+V|u|^n)\dd\leq1$, we prove the singular Moser–Trudinger inequality with sharp coefficient $(1-β/n)α_n$, where $α_n=n|\mathbb S^{n-1}|^{1/(n-1)}$. The subcritical supremum is attained for every such potential. If $V$ is radial and $r^βV(r)$ is nondecreasing, the critical supremum is also attained. This includes potentials that vanish on a ball and decrease to a positive limit at infinity. Under the same radial hypothesis, a bound for the tangential gradient shows that every extremal is radial and that the tangential gradients of maximizing sequences converge to zero in $L^n$, including at the critical coefficient. For the singular $n$-Laplacian equation with critical exponential growth, we obtain an explicit sufficient level for Palais–Smale compactness. Under the Ambrosetti–Rabinowitz condition and a smallness assumption at zero, this gives a nonzero nonnegative mountain-pass solution for all sufficiently large coupling parameters.

fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07628 ↗

We classify positive singular solutions of the boundary Yamabe equation on the punctured half-space in every dimension $n\ge3$. Previous symmetry results reduce this problem to a nonlinear PDE in two variables, for which classical phase-plane methods are unavailable. We develop a Hamiltonian method that combines local bifurcation with compactness to prove that every nonstationary cylindrical profile of a singular solution is periodic and uniquely determined, up to translation, by its Hamiltonian energy. At each stationary or periodic profile, the space of solutions of the linearized equation is two-dimensional. Finally, every positive solution in a half-ball with a nonremovable isolated boundary singularity is asymptotic to a unique global singular solution, with an algebraically decaying relative error.

comments
63 pages
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07636 ↗

We establish global uniqueness for the inverse boundary value problem of three-dimensional time-harmonic isotropic elasticity with smooth real Lamé parameters satisfying strong convexity and a smooth positive mass density. At a fixed known nonzero frequency for which the corresponding Dirichlet problems are uniquely solvable, the full displacement-to-traction map uniquely determines all three coefficients. The proof first recovers the full boundary Taylor expansions of $λ$, $μ$, and $ρ$, and derives an exact Green–Betti identity for a four-component reduction, with the terms involving $W=ω^2ρ$ and $\nabla W$ retained. Three independent physical polarizations of the rank-three characteristic quotient yield a complete $3\times3$ matrix of polynomial moment identities. A Cauchy transform construction produces an exact relative transport matrix $H$, normalized to $I_3$ in a common exterior region. Its invertibility follows from the determinant identity $\det H=\sqrt{μ_2/μ_1}>0$. Holomorphic dependence on the projective null direction, established by differentiating the transport equation at a fixed physical point, leads to a finite-type rigidity system that determines the Lamé parameters. Finally, products of physical elastic solutions and Fourier uniqueness recover the mass density.

fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07642 ↗

The present paper establishes the Richardson dispersion law, predicting the particle separation with time according to $\E R_t^2 \sim t^3$, as well as anomalous dissipation for the associated tracer, for a random Hölder continuous velocity vectorfield with controlled sweeping.

fields
Analysis of PDEs, Mathematical Physics, Probability
pdf ↗ page arXiv:2611.07694 ↗

We study radially symmetric solutions to the three-dimensional semilinear wave system with distinct propagation speeds \[ u_{tt}-c_1^2Δu=|v_t|^p\ , \ v_{tt}-c_2^2Δv=|u_t|^q\ , \ c_1, c_2>0\ , c_1\neq c_2\ . \] For \(1<p,q<2\), we prove that \[ Γ(p,q) = 2+\max\left\{\frac1p,\frac1q\right\}-pq=0 \] is the exact critical curve separating small-data global existence from finite-time blow-up. More precisely, sufficiently small, compactly supported radial initial data give rise to global solutions when $Γ(p,q)<0$. On the other hand, for a class of initial data, the corresponding solutions blow up in finite time, with an upper bound for the lifespan, when $Γ(p,q)\geq 0$.

comments
33 pages, 2 figures
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07710 ↗

For periodic initial data that allow vacuum, we establish the global-in-time existence of solutions to the 2D compressible nematic liquid crystal flows, under the condition that the bulk viscosity coefficient is sufficiently large and a geometric angular constraint is imposed on the initial direction field. Our results remain valid for arbitrarily large initial energy. Furthermore, under appropriate additional assumptions imposed on the pressure, we also establish the uniqueness of the constructed solutions.

fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07725 ↗

We prove global well-posedness for the focusing quintic nonlinear Schrödinger equation on the standard three-dimensional torus for initial data below the mass-corrected energy and Sobolev-norm thresholds determined by the Euclidean ground state. This gives a subthreshold focusing counterpart to the defocusing theorem of Ionescu and Pausader. The proof combines the periodic profile analysis with the recent nonradial Euclidean scattering theorem of Su and Zhao. The mass correction yields an invariant region and a nonnegative energy decomposition, placing every Euclidean profile below both ground-state thresholds. Uniform bounds on finite intervals follow from a contradiction argument on shrinking time intervals.

comments
14 pages. Comments are welcome!
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07740 ↗

We analyze the geometric variational crime of nonlinear finite element heterogeneous multiscale methods (FE–HMM) for $p$-growth monotone laws on curved surfaces, $p\ge2$. The practical method projects discrete gradients onto the exact tangent planes; we compare it with the exactly lifted method on the same discrete space. Without further hypotheses, the difference is $O(h_g^{p/(p-1)})$ in the metric induced by the cell correctors. The optimal order $O(h_g^2)$ holds with a constant given by a weighted negative moment of the corrected gradients, evaluated along the practical solution only; no bounds on discrete gradients in $L^\infty$ are used. The negative moment is a sufficient condition for the lower shifted $p$-structure of the discrete homogenized flux, which for $p>2$ is not available in general. For the model law and a fixed cell mesh, its uniform boundedness is equivalent to a finite-dimensional directional criterion on degenerate correctors, and for laminates it holds with an explicit bound, uniform in the cell mesh. With macro–micro quasi-optimality and an estimate for coefficient sampling, the reconstructed microscopic field converges as $O(h+H^{\min{s,1}}+h_g^2)$, where $H^s$ is the cell approximation rate. Experiments on a torus with an exact closest-point lift and a transported anisotropic $p=4$ material show second-order geometric convergence up to $98,304$ macroscopic quadrature points, each with a fully solved nonlinear cell problem, with bounded a posteriori constants for both load treatments.

fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07761 ↗

We prove that all known regular self-similar implosion solutions of the isentropic compressible Euler equations are unstable to shock formation. The perturbations can be chosen to be small in a class determined by the regularity threshold of the background. The proof relies on an amplification mechanism which occurs along the surface of the backward sonic cone from the implosion point. Motivated by our main results, we formulate an implosion censorship conjecture for the compressible Euler equations, stating that classical implosions are unstable to shock formation.

comments
40 pages, 7 figures
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07762 ↗

We consider the limit $δ\to1^+$ for the one-dimensional compressible Navier-Stokes equations with viscosity $μ(ρ)=ρ^δ$ and far field vacuum. To the best of our knowledge, this is the first result on the viscosity-exponent limit $δ\to1^+$ for strong solutions with far field vacuum. With far field vacuum, the diffusion coefficient $ρ^{δ-1}$ does not converge uniformly to $1$ on $\mathbb R$ as $δ\to1^+$. The main difficulty is to obtain uniform estimates for this limit. By introducing some new weighted density variables, we establish local well-posedness and uniform estimates without requiring $ρ_0^{(δ-1)/2}\in L^2(\mathbb R)$. We prove strong convergence to the known global $δ=1$ solution obtained in \cite{WenZhangDCDS2025} on a local time interval independent of $δ$. With one additional order of regularity of the initial data and the limit solution, the convergence holds on every fixed finite time interval, and the maximal life-spans of the $δ>1$ solutions tend to $+\infty$ as $δ\to1^+$.

fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07790 ↗

We develop a unified Lagrangian framework for the uniqueness of weak solutions to a broad class of two-dimensional inviscid active scalar equations in generalized mean oscillation spaces. Under common structural assumptions on the Biot–Savart kernels, we identify mean oscillation spaces that ensure an Osgood condition for the velocity, and establish uniqueness in these classes. For the two-dimensional Euler equations, we establish uniqueness for vorticities whose mean oscillation grows mildly at small scales, extending beyond $\mathrm{BMO}$. For the surface quasi-geostrophic (SQG) and incompressible porous media (IPM) equations, we obtain uniqueness in spaces logarithmically rougher than $C^{0,1}$. We also obtain what are (to our knowledge) the strongest known uniqueness results for the $α$-SQG equation, $0<α<1$, in spaces logarithmically rougher than critical Hölder spaces $C^α$. The framework also applies to various logarithmically perturbed models, yielding uniqueness in the corresponding sharp regularity classes. The proof combines a sharp regularity estimate for the velocity with a stability estimate for integral operators under measure-preserving rearrangements. This provides a single uniqueness argument for the entire family under common structural assumptions on the kernels.

comments
33 pages
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07834 ↗

We develop a continuum model to quantify kinematic collision risk in stationary crowd flow. From the density-weighted inward relative flux, we derive a collision functional based on the symmetric velocity gradient. For incompressible flows with constant density, this functional is convex and positively homogeneous of degree one in a combination of second derivatives of the stream function. We minimize it subject to prescribed boundary flux. On bounded simply connected domains with Lipschitz boundary, under a finite-energy lifting assumption, we prove existence of minimizers in a relaxed class that allows this derivative combination to be a vector-valued Radon measure. We show that disks are the only domains with a nontrivial zero-trace kernel, corresponding to rigid rotation about the center. Using the compact embedding for functions of bounded deformation, we prove that relaxed energy sublevels are compact without normalization on every other admissible domain, and after removing one rotational component on disks. We identify the relaxed functional with the strong $W^{1,1}$ lower semicontinuous envelope of the functional on the prescribed-trace class with integrable controlled derivatives. Then, we derive optimality conditions via convex duality. We construct explicit radial flows on annular sectors and prove that they minimize the energy over the full admissible classes, including nonradial competitors. On the square, distinct minimizers with the same prescribed boundary flux have different local collision intensities. The square–disk comparison shows that the saddle flow loses optimality when the domain changes, even after normalization.

comments
28 pages, 6 figures
fields
Analysis of PDEs, Optimization and Control
pdf ↗ page arXiv:2611.07837 ↗

We study blow-up solutions of generalized Liouville systems on a closed surface at and near critical parameter hypersurfaces. Under a Morse assumption on the reduced functional, we establish the following results. First, we introduce amplitude variables in the approximate manifold and prove that every sequence of blow-up solutions at a fixed critical parameter, or along an admissible sequence from the lower side, admits a unique minimizing decomposition up to permutation. The remainder converges to zero and is orthogonal to the center, scale, and amplitude directions. The concentration scales are comparable, the limiting configuration satisfies the scale balance conditions, and the amplitudes vanish after multiplication by the logarithmic scale. Second, we determine the leading parameter defect in both critical decay regimes and prove that, along each fixed admissible path, all sufficiently large solutions lie in finitely many disjoint bubbling neighborhoods. Finally, assuming the leading coefficient is nonzero at every balanced critical configuration, the solution set at the fixed critical parameter is compact. These results provide the analytic foundation for studying the resonant degree via critical points at infinity.

comments
44 pages
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07857 ↗

In this work, we present global solvability results, in critical Sobolev spaces, for divergence-type equations associated with elliptic systems of vector fields.

fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07904 ↗

In this paper, we study the first Dirichlet eigenfunction on bounded, smooth and strictly horo-convex domains in hyperbolic space $\mathbb{H}^n$. We prove its strict $\log$-concavity in arbitrary dimension and establish the Gauss curvature bound $K>1$ for its level sets. In dimensions two and three, we further obtain quantitative lower bounds for $K$ in terms of the minimum Gauss curvature of the boundary. In higher dimensions, the product of the two smallest principal curvatures of each level hypersurface exceeds one. Furthermore, we extend the strict $\log$-concavity result to bounded smooth horo-convex domains. Our approach combines maximum or minimum principles for suitable auxiliary functions, boundary convexity estimates and a domain deformation argument.

comments
51 pages
fields
Analysis of PDEs, Differential Geometry
pdf ↗ page arXiv:2611.07907 ↗

We consider the Minkowski mean curvature equation $$ -\operatorname{div}\left(\frac{\nabla u}{\sqrt{1-|\nabla u|^2}}\right) =λu+μh(x,u) \quad \text{in } Ω, \quad \mathcal B u = 0 \quad \text{on } \partial Ω, $$ where $\mathcal B \in \{ \mathcal D, \mathcal N \}$ is a Dirichlet or Neumann boundary condition and $Ω$ is a smooth bounded domain, required to be convex in the Neumann case. Moreover, $λ, μ\geq0$ are parameters, and $h(x,u)$ is odd with respect to $u$ and positive for $u > 0$. Using minmax methods in nonsmooth critical point theory, we prove that $λ>λ^\mathcal B_k$ yields at least $k$ pairs of nontrivial solutions with negative energy for every $μ\geq0$, where $λ_k^\mathcal B$ are the corresponding Laplacian eigenvalues, indexed by $λ^\mathcal B_0=0$. Moreover, if $H(x,u):=\int_0^u h(x,t)\,dt=o(u^2)$ as $u\to0$, we prove that for $λ\in[λ^\mathcal B_k,λ^\mathcal B_{k+1})$ and every $j\geq\max\{1,k\}$, sufficiently large $μ$ yields at least $2j-k$ distinct pairs of solutions: $j$ with negative energy and $j-k$ with positive energy.

fields
Analysis of PDEs
pdf ↗ page arXiv:2611.07963 ↗

We study the Boltzmann equation for a gas on the three-dimensional torus. Its particles collide with one another and with a thermal reservoir whose temperature varies in space. The reservoir heats the gas in some regions and cools it in others, driving a nonequilibrium energy current. For the isotropic cutoff kernel, with Maxwell molecules or hard potentials, we prove global well-posedness and exponential convergence to the same stationary state for initial data with a common Gaussian upper bound and Gaussian moment bound. These bounds determine how weak self collisions must be relative to bath collisions. The estimates are uniform over bath profiles between fixed positive temperature bounds. For any nonconstant reservoir profile on the three-dimensional torus, this stationary state cannot be a local Maxwellian. When particles collide only with the bath, relative entropy to the stationary state cannot increase. Once self collisions are added, arbitrarily small perturbations can initially increase it, even though they later return exponentially.

fields
Analysis of PDEs, Mathematical Physics
pdf ↗ page arXiv:2611.07999 ↗

In this work, we analyze the initial boundary value (IBV) problems for the Short Pulse (SP) equation $u_{xt}=u+\frac{1}{6}(u^3)_{xx}$ (classical solutions) on the half-lines $x\ge 0$ and $x\le 0$ under the condition of sufficient decay as $x\to +\infty$ or as $x\to -\infty$. Using the Riemann–Hilbert (RH) problem approach, we show that these two IBV problems differ substantially regarding the boundary values required for problems to be well-posed. Namely, the solution of the ``right problem'' ($x\ge 0$) is uniquely determined by the initial data $u(x,0)$, $x\ge 0$ alone, whereas for the ``left problem'' ($x\le 0$), two boundary functions $u(0,t)$ and $u_x(0,t)$ have to be prescribed, together with the initial data, for a well-posed problem.

comments
53 pages, 4 figures
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.08105 ↗

We study the critical Choquard equation \[ -Δu-μ|x|^{-2}u =\bigl(|\cdot|^{2-N}*|u|^p\bigr)|u|^{p-2}u \qquad\text{in }\mathbb R^N, \] where $|\cdot|^{2-N}$ is the Newton kernel, $p=\frac{N+2}{N-2}$, and $0<μ<\frac{(N-2)^2}{4}$. For every $N\ge3$, we prove that positive solutions in $D^{1,2}(\mathbb R^N)$ are radially symmetric about the origin, unique up to dilation, and nondegenerate. Our proof uses normalized logarithmic derivatives to obtain a cooperative system. The resulting monotonicity principle underlies both the uniqueness and nondegeneracy arguments, and remains effective in the subquadratic range $1<p<2$, corresponding to $N\ge7$. Consequently, the linearized quadratic form has Morse index one and the sharp inequality admits a global quadratic remainder estimate. The best constant is strictly concave in the Hardy parameter on $\bigl[0,\frac{(N-2)^2}{4}\bigr)$, is of class $C^2$ with strictly negative second derivative in the interior, and satisfies two-sided bounds that agree to first order at $μ=0$.

comments
37 pages, comments are welcome
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.08146 ↗

We study the Boltzmann equation for a non-isothermal plane Couette flow in the three-dimensional channel $\mathbb T_x\times(-1,1)_y\times\mathbb T_z$ for Maxwell molecules. For sufficiently small shear rate and wall temperature perturbations, we construct a unique steady solution near the global Maxwellian. For the corresponding unsteady boundary value problem, we establish the global existence and uniqueness of solutions arising from small perturbations of the steady state, together with their exponential convergence to the steady state as $t\to\infty$. A major difficulty in both the steady and unsteady problems is to close the outgoing mass fluxes generated by diffuse reflection. To overcome this difficulty, we develop a new boundary flux transmission framework which, in the slab geometry, yields steady and time-dependent wall flux estimates through the boundary-to-boundary transmission between the two walls. We further extend this framework to damped transport equations with diffuse reflection in general bounded $C^1$ domains, even when the damping coefficient degenerates at large velocities.

comments
61 pages
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.08159 ↗

For $2\le k<n$, we prove pointwise gradient bounds for proper weak $k$-convex functions in terms of a local average and a Wolff potential of the full Hessian measure $F_k[u]$. The estimate yields local Lipschitz regularity for $F_k[u]=f$ with $f\in L^{n/k,1/k}_{\mathrm{loc}}$, and for $F_k[u]=|Du|^p$ when $0<p<nk/(n-k)$. Both the Lorentz index $1/k$ and the upper bound on $p$ are sharp.

fields
Analysis of PDEs
pdf ↗ page arXiv:2611.08184 ↗

We create counterexamples to show that Pólya-Szegő type of inequality fails for the weighted nonlocal seminorm \begin{equation*} [u]_{W^{s,p, α}(\mathbb{R}^d)} := \left( \int_{\mathbb{R}^d} \int_{\mathbb{R}^d} \frac{|x|^{α_{1} p} |y|^{α_{2} p} |u(x)-u(y)|^{p}}{|x-y|^{d+sp}} \, dx \, dy \right)^{\frac{1}{p}} \ , \end{equation*} in all three regimes $α_1 +α_2 < 0, \, α_1 +α_2 = 0$, and $α_1 +α_2 >0$. These counterexamples are such that it completely settles the first two cases for all parameters. We also provide necessary and sufficient conditions on the parameters for $ [u]_{W^{s,p, α}(\mathbb{R}^d)} < \infty$, thus extending an earlier known result of [Serena Dipierro and Enrico Valdinoci, A density property for fractional weighted Sobolev spaces]. We also provide another proof of the fractional C-K-N type inequality, for most of the parameters (Hardy and Sobolev are included), using a familiar technique which is involved in the proof of fractional Sobolev inequality of [Eleonora Di Nezza, Giampiero Palatucci, and Enrico Valdinoci, Hitchhiker's guide to the fractional Sobolev spaces, Bulletin des sciences mathematiques] together with a symmetrization technique.

comments
26 pages, 1 figure
fields
Analysis of PDEs, Functional Analysis
pdf ↗ page arXiv:2611.08189 ↗

We analyse stationary fully nonlinear uniformly elliptic second-order mean field games with nonseparable and nondifferentiable Hamiltonians by introducing the notion of viscosity-variational inequality solution. In this formulation of the problem, the value function is the viscosity solution of the Hamilton–Jacobi–Bellman (HJB) equation, and the Kolmogorov–Fokker–Planck (KFP) equation for the player density is understood as a nonstandard variational inequality that no longer involves partial derivatives of the value function. The advantages of this formulation are found in its stability and continuity properties for handling weak convergence arguments, which we use to prove the existence of solutions of the problem. Furthermore, we prove the equivalence of the notion of viscosity-variational inequality solution with the notion of a partial differential inclusion solution, where the value function is a strong solution of the HJB equation and the player density is a very weak solution of the KFP equation with coefficients belonging to subdifferential sets of the Hamiltonian. Finally, we prove the uniqueness of solutions in the case of separable Hamiltonians with strictly monotone couplings.

comments
38 pages
fields
Analysis of PDEs, Optimization and Control
pdf ↗ page arXiv:2611.08201 ↗

Let $n \ge 3$, $2 \le α< n$, and $1 < p,r < \infty$ satisfy $p^{-1}+r^{-1}=1+α/n$. We prove uniqueness, up to common translations and dilations, of the normalized nonnegative extremizing pairs for the corresponding Hardy-Littlewood-Sobolev inequality. More generally, all positive finite-energy solutions of the Euler-Lagrange system belong to a single symmetry orbit. The proof compares two solutions at different relative scales. Total positivity of order two identifies the second eigenspace of an associated compact self-adjoint operator and forces each nonzero difference to change sign exactly once. The opposite orientations obtained at the two ends of the scale parameter then yield uniqueness. The same method proves radial nondegeneracy modulo dilation. We also establish strong convergence of the normalized alternating Euler-Lagrange iteration from every nonnegative, radially nonincreasing initial datum to a fixed extremizing pair, without rescaling the iterates. Square-root mass variables reduce the exclusion of several concentration profiles to a static spectral estimate and an energy-increment argument. The analysis includes sublinear exponents and the logarithmic borderline decay regimes.

fields
Analysis of PDEs, Classical Analysis and ODEs
pdf ↗ page arXiv:2611.08222 ↗

This paper establishes the global well-posedness of shock solutions for a 3D spherical piston problem to the compressible Euler equations. The main mathematical challenges are the nonlinear coupling between the free shock boundary and the moving piston, alongside potential gradient catastrophes induced by spherical geometric source terms. To overcome these difficulties, we reformulate the problem using Riemann invariants and introduce a tailored directional derivative along the unperturbed piston trajectory. Then we can extend the local solution to construct a global piecewise smooth solution including a single shock under some moderate assumptions.

fields
Analysis of PDEs
pdf ↗ page arXiv:2611.08260 ↗

We establish quantitative local higher integrability estimates for weak solutions of degenerate parabolic systems with generalized double phase growth, whose prototype is $$u_t-\operatorname{div}\left(\frac{G'(|Du|)}{|Du|}Du+a(x,t)\frac{H'(|Du|)}{|Du|}Du\right)=0\qquad\text{in }Ω_T.$$ Here $G$ and $H$ are Young functions satisfying suitable structural assumptions. The nonnegative coefficient $a$ is allowed to vanish and has a concave modulus of continuity $ω$ with respect to the parabolic distance. For $Ψ(z,\varkappa)=G(\varkappa)+a(z)H(\varkappa)$, we prove that solutions in $L^\infty(0,T;L^s(Ω,\mathbb{R}^N))$, $2\le s<\infty$, satisfy $Ψ(\cdot,|Du|)\in L^{1+\varepsilon}_{\mathrm{loc}}$ for some $\varepsilon>0$, provided $$\limsup_{ρ\to 0+}ω(ρ)\frac{H(ρ^{-\frac{n+s}{s}})}{G(ρ^{-\frac{n+s}{s}})}<\infty.$$ For bounded solutions, the corresponding condition involves $H(ρ^{-1})/G(ρ^{-1})$. These conditions extend the known interpolative gap bounds from power growth to general growth functions and continuity moduli. In particular, for $G(\varkappa)=\varkappa^p$, $H(\varkappa)=\varkappa^q$, and $ω(ρ)\simeqρ^α$, they recover $q\le p+sα/(n+s)$ and $q\le p+α$, respectively. The proof combines an intrinsic stopping-time argument with inverse-function estimates to derive Poincaré and reverse Hölder inequalities adapted to the general growth phases.

comments
54 pages. Comments are welcome
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.08291 ↗

We consider the so-called Weighted Energy-Dissipation approximation of gradient flows in metric spaces. We prove the convergence of minimizers to maximal slope curves under assumptions that are close to those of the original existence theory for metric gradient flows. The proof relies on an energy-dissipation identity for the approximating curves and on an adaptation of a general stability result for maximal slope curves.

fields
Analysis of PDEs
pdf ↗ page arXiv:2611.08377 ↗

Optimization and Control39

The complexity of finding second-order stationary points (SOSPs) in constrained settings has remained unclear and has been highlighted as an open question by the works of Hollender and Zampetakis [2023] and Kontogiannis et al. [2024]. In particular, under one established definition (SOSP1) where nonnegative curvature is required only along the null space of the active constraints [Bertsekas, 1997], the problem lies in TFNP, and algorithms with $O(poly(1/ε))$ running times have been proposed. Under another established notion of stationarity (SOSP2) checking is coNP-hard in high dimensions. We investigate the computational landscape of finding SOSPs under polytope constraints. We prove that, even if promised that all KKT points lie a constant distance away from the domain's boundary–where SOSP1 and SOSP2 reduce to the standard unconstrained conditions–finding an approximate SOSP under either definition is PLS-hard, and the hardness continues to hold even in $[0,1]^2$. As a direct consequence, our result establishes a fundamental barrier on the continuity of SOSP-computing algorithms: unless PLS=CLS, no deterministic, iterative algorithm with a computationally efficient, continuous update rule can guarantee that every $ε$-approximate fixed point is an $(ε,\sqrtε)$-SOSP, even under the above interior promise. We establish PLS-completeness by proving that the problem of finding approximate SOSP1 belongs to PLS in high dimensions. In addition, we prove that, for every fixed $d\ge 2$, finding an approximate SOSP2 is also PLS-complete, escaping the coNP-hardness which appears in high dimensions. We also establish PLS-hardness and a polynomial-time search reduction to LOCALOPT for finding $(r^{2+γ},r)$-local pure Nash equilibria, for every fixed rational $γ\in(0,1]$, in two-player nonconvex common-interest games with actions in $[0,1]$.

comments
This version simplifies the group analysis and extends the results to the SOSP2 definition and nonconvex games. The abstract was shortened to meet the arXiv requirements
fields
Optimization and Control
pdf ↗ page arXiv:2604.02285 ↗

This paper establishes a rigorous connection between Noether's second theorem and the structure of Pareto fronts in multi-criteria variational problems. We prove that for a family of functionals invariant under a common gauge Lie group, the regular part of the Pareto front is contained in the union of solutions to weighted Euler–Lagrange equations, and that on Pareto-optimal sections, combinations of Noether currents are conserved. Gauge symmetries are shown to impose additional constraints on the Lagrange multipliers via Bianchi-type identities, reducing the dimension of the admissible weight space. Two practical approaches are derived from this result: dimensionality reduction via gauge fixing and scalarization with penalty terms. The theory is validated on three classical problems: the linear-quadratic regulator with two criteria, a model predictive control problem with a minimax criterion, and the double integrator with control constraints. In each case, the Pareto front is shown to satisfy an exact conservation law, and the analytical structure of the front is recovered from the symmetries of the problem. The results demonstrate that gauge functions (null Lagrangians) can restore invariants even when the dynamics break the symmetry, and that control constraints do not destroy the underlying conservation laws but merely restrict the admissible set. The proposed approach yields a Pareto front with minimal deviation from the theoretical curve, outperforming known multi-objective algorithms, and provides a new foundation for the analytical treatment of multi-criteria optimization in nonlinear dynamics and control.

comments
24 pages, 4 figures
fields
Optimization and Control, Mathematical Physics
pdf ↗ page arXiv:2611.06920 ↗

The local convergence of primal-dual interior-point methods for semi-definite optimization is closely tied to the regularity of the central path at its limiting point. Without strict complementarity, the central path can exhibit fractional power asymptotics and need not be differentiable in the ordinary barrier parameter. We show that this obstruction can be removed simultaneously for a family of weighted \AHO~central paths. More precisely, there exists a common ramification exponent $N$ such that all weighted paths in a fixed neighborhood of the identity become analytic at the limiting point after the reparametrization $μ\mapstoμ^N$. Using higher-order derivatives of these reparametrized paths, we construct a feasible $p$th-order predictor-corrector interior-point method. Under a uniform predictor regularity condition, and without assuming strict complementarity, we prove that the normalized complementarity gaps converge with $Q$-order at least $(p+1)/ρ$, where $ρ$ is the reparametrization exponent used by the method. We also establish super-linear convergence of the simultaneously vanishing primal-dual blocks, derive an explicit quantitative reduction bound throughout the predictor regularity region, and obtain a doubly logarithmic outer iteration bound once the iterates enter the local super-linear regime. Finally, we illustrate the predicted higher-order behavior on a collection of degenerate and non-degenerate semi-definite optimization problems.

fields
Optimization and Control, Algebraic Geometry
pdf ↗ page arXiv:2611.07080 ↗

As the importance of semialgebraic optimization grows in engineering, so does the need for scalable sum-of-squares (SOS) methods. Their convex relaxations can guide mixed-integer search and often deliver certified global optima outright. Scaling traditional SOS typically requires hand-crafted bases and redundant constraints, whereas sample-based SOS numerically exploits feasible set geometry to reduce bases and build smaller, often tighter relaxations from black-box evaluations. However, discrete variables can require exponentially many samples, and each sampled equation couples all cliques, destroying correlative sparsity. In this work, we avoid enumeration and slack variables by sampling a tractable variety that contains the feasible set, with multipliers only for the omitted constraints. To recover sparsity, we introduce "separator" polynomials that split the sampled equations into local systems coupled only on clique overlaps. This yields a novel, convergent hierarchy of semidefinite programs (SDPs) with block sizes and sample requirements governed by local interactions rather than total problem size. Lastly, we show how to verify sample sufficiency and prune redundant equations numerically before solving. In robust estimation and motion planning, we obtain certifiably small optimality gaps at scales far beyond those reported for traditional sparse SOS. For example, we solve a truncated least-squares 3D registration problem with 1600 candidate correspondences in under 49 seconds using an off-the-shelf SDP solver on a laptop CPU.

comments
25 pages, 4 graphics. To be submitted to SIOPT. Codebase: https://github.com/Rhiza-PolyOpt
fields
Optimization and Control
pdf ↗ page arXiv:2611.07108 ↗

On July 23, 2026, at the opening of the International Congress of Mathematicians in Philadelphia, the International Mathematical Union announced the award of the Gauss Prize to Yurii Evgenievich Nesterov -- an award for mathematical work with significant applications outside mathematics. This article explains the mathematical content of the recognized work of Yu. E. Nesterov and traces its influence on modern computational practice. The exposition is based on a concept that Yu. E. Nesterov himself regards as one of his main results: the classical complexity theory of convex optimization is oriented toward a black-box oracle, but its pessimistic bounds can be overcome by using a priori information about the internal structure of the objective function. We trace Yu. E. Nesterov's scientific career, then introduce the language of oracle complexity together with a classification of objective function models (geometric, algebraic, and structural), and subsequently discuss his main results in chronological order: the fast gradient method, fast automatic differentiation, the theory of self-concordant barriers and polynomial-time interior-point methods, the $π/2$ theorem, the stable dynamics model of traffic flows, the smoothing technique, primal-dual methods, cubic regularization of Newton's method, the modified Gauss–Newton scheme, coordinate descent methods and methods for huge-scale problems, gradient-free methods, universal gradient methods, implementable tensor methods, and "superfast" methods. Each topic is accompanied by historical background, precise statements specifying the problem class, oracle, and type of guarantee, examples, and a discussion of the current state of the art.

comments
115 pages, in English; 7 figures, 4 tables, 191 references
fields
Optimization and Control
pdf ↗ page arXiv:2611.07114 ↗

Distance based formation control fixes the shape of a robot team but leaves its position and orientation unconstrained. We regulate all three for differential drive robots that have only one absolute position anchor, absolute headings, body frame relative position measurements, and communication with their neighbors. We derive a two layer estimation control scheme directly from optimization. The first half of the result is a set of graph-proximal estimators that reconstruct the centroid, the anchor vector, and the centered coordinates. Exact proximal updates reach consensus in finitely many steps, and a finite number of primal dual iterations leaves a bounded reconstruction error. The controller combines proximal centroid and orientation steps with distance gradient shape regulation, and a finite rotation removes the orientation drift caused by the shape step. A discrete Lyapunov analysis gives local exponential convergence once the estimators have converged. We validate the method in ROS 2 Gazebo and in experiments with five physical robots.

comments
11 pages, 6 figures
fields
Optimization and Control
pdf ↗ page arXiv:2611.07163 ↗

Stopping games are the multiplayer generalization of optimal stopping problems, in which each player independently chooses when to stop, and the payoff depends on who stops first and when. While the existence of $\varepsilon$-equilibria in two-player stopping games with integrable payoffs is known, the multiplayer case has remained open. We prove that every multiplayer stopping game in discrete time with integrable payoff process admits an $\varepsilon$-equilibrium for every $\varepsilon > 0$, provided the filtration is conditionally atomless. This condition is satisfied, in particular, when the underlying probability space is $\{0,1\}^{\dN}$ endowed with the uniform measure and its natural filtration. Our method also shows that every three-player stopping game, without the restriction to conditionally atomless filtration, admits an $\varepsilon$-equilibrium, for every $\varepsilon > 0$. The proof combines techniques developed for two-player stopping games and for multiplayer quitting games, with a key new insight: the richness encoded in a conditionally atomless filtration can substitute for an external public randomization device, allowing players to coordinate without an explicit coordination signal.

fields
Optimization and Control, Probability
pdf ↗ page arXiv:2611.07184 ↗

We prove that deciding whether an integer cubic polynomial takes a value at most one over three nonnegative integers is NP-hard. Hardness holds even when every feasible objective value is a positive integer, so every constructed instance has a finite, attained minimum. This establishes a sharp dimension boundary for the tractability of cubic integer optimization in fixed dimension. Our reduction from positive 1-in-3 satisfiability encodes Boolean constraints as modular conditions and uses an algebraic norm to construct a cubic objective whose minimum is one exactly for satisfiable formulas. The resulting six-parameter threshold language is NP-complete and admits short exponent certificates.

comments
16 pages. A version of this manuscript was posted on Optimization Online on November 1, 2026
fields
Optimization and Control
pdf ↗ page arXiv:2611.07269 ↗

A team of unmanned aerial vehicles (pursuers) with different positions, headings, speeds, minimum turning radii and launch delays must respond to a target that moves at constant velocity through a region containing static circular obstacles. Only one pursuer is launched, namely the one that can intercept the target first. We present a computationally efficient, obstacle-aware interceptor-selection framework for this setting. The interception problem of each pursuer is posed as a minimum-time optimal control problem and is solved over an explicit family of candidate path types through an arc-parameterised nonlinear programme in which the obstacles are toured tangentially and every part of the path is kept out of every obstacle by exact (not sampled) distance constraints. A closed-form lower bound, obtained by relaxing the curvature and obstacle constraints, is combined with the verified feasible interception times in a bound-and-select algorithm. We prove that the pruning is certified: a pursuer is skipped only if it cannot improve the best verified interception time, so bound-and-select returns the same pursuer and the same time as solving every pursuer, and the method reports a boundary on the optimal team interception time. Numerical experiments with solver diagnostics and repeated random starts, sensitivity studies in the target speed (with the selection crossover located by bisection), the launch delay, the turning radius and the obstacle configuration, a comparison with nearest-pursuer, fastest-pursuer and straight-line-time selection rules on random scenarios, a computational comparison with full enumeration for growing team sizes, and a geographic case study show that the selected pursuer is frequently neither the nearest nor the fastest, and that lower-bound pruning removes a large part of the computational effort without changing the decision.

comments
29 pages and 4 figures
fields
Optimization and Control
pdf ↗ page arXiv:2611.07279 ↗

Last-iterate convergence guarantees for stochastic gradient methods have attracted significant attention as practical implementations typically return the last iterate instead of an average of the iterates. In this paper, we study two approaches for analyzing the last-iterate convergence of proximal and projected stochastic gradient methods without bounded domain or bounded variance assumptions, where black-box variational inference serves as a running example. Under two related assumptions relaxing the bounded variance, we show near-optimal anytime rate $\widetilde{\mathcal{O}}(1/\sqrt{T})$ for the last iterate. Moreover, when the horizon $T$ is known, we remove the logarithmic factor and obtain the optimal order $\mathcal{O}(1/\sqrt{T})$, matching the order known under bounded variance or bounded domain assumptions.

comments
31 pages
fields
Optimization and Control
pdf ↗ page arXiv:2611.07334 ↗

Parameter-efficient fine-tuning with LoRA is governed by rank, which controls parameter budget and locally accessible task updates. Existing dynamic-rank methods mainly focus on capacity allocation, leaving unresolved whether the current adaptation space is sufficient for the task. We characterize the first-order information boundary of standard factor gradients, showing that their unobservable subspace coincides with the normal space of the low-rank manifold. Thus, factor gradients alone cannot distinguish under-optimization within the current space from task-relevant directions beyond its first-order reach. This motivates a geometric view of rank evolution, where tangent information optimizes existing degrees of freedom and the normal spectrum quantifies the value of new ones. Under a local quadratic model, we show that leading normal singular directions yield the locally optimal rank expansion, with spectral energy determining its gain. Building on this characterization, we propose GEAR, which turns the tangent–normal decomposition into rank evolution through tangent-space optimization, matrix-free normal-spectrum estimation for rank expansion, and task-induced geometric initialization. Across language and vision benchmarks, GEAR achieves the best average performance in all evaluated settings with substantially fewer trainable parameters. On LLaMA3-8B, it raises the average commonsense score from 86.61 to 88.45 while reducing trainable parameters from 42.0M to 14.4M.

fields
Optimization and Control
pdf ↗ page arXiv:2611.07349 ↗

We develop a structured continuous-time q-learning framework for finite-horizon investment and optimal stopping under jump-diffusion dynamics with unknown parameters. A penalty approximation transforms the Hamilton-Jacobi-Bellman variational inequality into an HJB equation with a binary control for the stopping penalty. An auxiliary state incorporates the resulting control-dependent discount factor into a Markovian formulation. We randomize controls on the mixed action space and introduce Shannon entropy regularization, establishing policy improvement and martingale characterizations of the continuous-time q-function. These results motivate a structured q-learning algorithm that constructs the q-function from the penalty HJB equation and jointly learns the value function and unknown generator parameters from sample paths. We apply the method to portfolio choice with optimal stopping in a simulation study. We further conduct a two-asset empirical study, evaluating learned strategies on bootstrap paths from held-out test returns versus an MLE-finite difference benchmark.

fields
Optimization and Control
pdf ↗ page arXiv:2611.07362 ↗

Stochastic Gradient Descent (SGD) is one of the most classical optimization algorithms with favorable theoretical guarantees, yet its practical implementation differs subtly from its well-known form and is often referred to as Shuffling Stochastic Gradient Descent (Shuffling SGD). A particularly popular strategy in Shuffling SGD is Random Reshuffling (RR), which has achieved great empirical success. Despite its strong performance, RR has long been considered a heuristic due to a lack of theoretical support. Over the last decade, people have finally established provable convergence rates for RR, thus justifying its observed superiority. However, for smooth convex optimization, two clouds over the convergence theory of RR remain. According to current theory, Shuffling SGD under RR converges only when the stepsize is smaller than a threshold proportional to $1/n$, where $n$ is the number of summands in the objective (or the number of data points). Consequently, the optimally tuned theoretical rate of Shuffling SGD under RR is strictly worse than that of SGD when the number of epochs is smaller than another threshold proportional to $n$. These two restrictions heavily limit the applicability of existing theories and leave a critical mismatch with practice. In this work, for the first time, we prove that RR dominates SGD in smooth convex optimization for any reasonable stepsize after any finite number of epochs, thereby addressing a longstanding open question. Moreover, we establish new algorithm-dependent lower bounds for RR and SGD, respectively. Altogether, these results yield (nearly) sharp worst-case characterizations of RR and SGD. Furthermore, we develop a shuffling-based method to optimize functions satisfying the quadratic growth condition, a strict relaxation of strong convexity, and derive a new complexity bound improving existing results proved under stronger assumptions.

comments
Extended version of arXiv:2606.32005. Originally submitted to arXiv on September 29, 2026
fields
Optimization and Control
pdf ↗ page arXiv:2611.07406 ↗

Truncated backpropagation through time reduces the computational burden of training recurrent models by restricting gradient computations to shorter temporal contexts. This induces two distinct boundary effects: missing past information perturbs the forward state evolution, while truncating the backward recursion omits future information. This motivates decoupling the interval over which gradient contributions are averaged for the parameter update from the surrounding temporal context over which they are computed. Under exponential decay of the state-transition Jacobians and mild regularity conditions, this leads to a two-sided gradient turnpike property: the error between truncated- and full-context gradient contributions decays exponentially with the distance from both context boundaries. We derive bounds for the averaged update gradient, yielding guarantees on gradient alignment and approximate full-context stationarity; under unbiased interval sampling and a Polyak-Lojasiewicz condition, these further imply a bound on the resulting optimality gap. Numerical experiments on the Wiener-Hammerstein benchmark illustrate the role of past and future context and the trade-off between predictive performance and computational cost.

fields
Optimization and Control
pdf ↗ page arXiv:2611.07415 ↗

In many modern multi-agent systems, a mediator seeks to influence strategic decision dynamics using only outcome-level feedback and without knowing agents' utilities. In this work, we study how a mediator can learn an unknown game and steer adaptive agents toward outcomes that maximize her utility using only trajectory feedback. The mediator knows the game form and her own objective, but not the agents' utilities. Moreover, she commits to nonnegative payments and provides recommendations and contexts, but receives only a single realized terminal outcome rather than full-information feedback. We develop halving-type algorithms that learn utility differences in normal-form games and conditional continuation-value differences in extensive-form games, achieving a sample complexity of $\widetilde O(\varepsilon^{-2})$ that matches worst-case lower bounds up to logarithmic factors. We then combine these learning procedures with a Lagrangian formulation in an online Learn-Compute-Steer algorithm that interleaves learning, equilibrium computation, and steering toward mediator-optimal extensive-form correlated equilibria (EFCE). With high probability, the Learn-Compute-Steer algorithm achieves a directness gap and an optimality gap of $\widetilde O(T^{-1/4})$, while time-average payments are of $\widetilde O(T^{-1/8})$.

comments
57 pages, 5 figures
fields
Optimization and Control
pdf ↗ page arXiv:2611.07589 ↗

This paper proposes a bilevel particle swarm optimization (PSO)-neurodynamic method for solving the distributed sparse least-squares problem regularized by the smoothly clipped absolute deviation (SCAD) penalty to global optimality. The SCAD penalty is nonsmooth and nonconvex, so existing distributed neurodynamic methods have not been extended to it directly. Through two-time-scale singular perturbation analysis, a rigorous full-trajectory convergence theory with explicit convergence rates is established for the inner collaborative multi-agent neurodynamic system without requiring the second-order sufficient condition. A theory-guided inner-outer collaborative optimization framework is further developed, in which a red-light green-light sensitivity mechanism is embedded in the outer PSO search to enhance global search while concentrating computational effort on the SCAD shape parameters, thereby improving recovery accuracy and computational efficiency. Simulation results on four sparse-recovery scenarios demonstrate that the proposed method accurately recovers the true sparsity pattern and achieves faster convergence and lower relative error than competing neurodynamic methods.

fields
Optimization and Control
pdf ↗ page arXiv:2611.07618 ↗

This paper proposes a continuous-time, projection-free neurodynamic system based on gradient tracking for distributed nonnegative sparse optimization. Unlike traditional methods, our approach avoids projecting states onto the feasible set. Instead, it employs gradient tracking to estimate average gradients under heterogeneous costs, coupling primal and dual dynamics such that feasibility and sparsity emerge naturally from the equilibrium conditions. The system ensures unique solutions through a locally Lipschitz vector field and guarantees that every equilibrium corresponds to an exact KKT point. We establish rigorous convergence guarantees for convex objectives through three distinct analytical layers. These include a centralized kernel rate, a computable distributed rate derived from Jacobian analysis without restrictive stability assumptions, and a global exponential stability proof based on Lur'e system theory. Our analysis quantitatively separates these layers, revealing that the convergence rate is primarily limited by graph sparsity and remains robust to variations in momentum time scales. Numerical experiments on strongly convex quadratic problems validate the theoretical rates, demonstrate effective active set recovery, and show that the system maintains stable tail convergence rates even under strong local heterogeneity.

fields
Optimization and Control
pdf ↗ page arXiv:2611.07643 ↗

We develop a graph-theoretic approach to the expected-norm matrix arithmetic–geometric mean inequality. By encoding matrix interactions with a path-supported Gram matrix, we obtain a projection family whose product is nonzero exactly when the transition graph of its index sequence is a path. This gives an intuitive explanation for failure: repeated indices can create cycles or branches that annihilate products, an effect absent for positive scalars. We prove the universal lower bound $(n)_m/n^m$ for the ratio of the with-replacement to the without-replacement mean and show that it is asymptotically sharp on the exponential scale when $m=n$. Specifically, our construction gives a ratio $\exp(-n+O(\sqrt n))$, matching the lower bound's leading exponential rate over unrestricted dimensions. We also determine an explicit overlap threshold, derive exact six-factor counterexamples, and show that exponential failure can coexist with vanishing pairwise commutators.

fields
Optimization and Control
pdf ↗ page arXiv:2611.07649 ↗

We study the minimum number of query rounds needed to find approximate stationary points of nonsmooth weakly convex functions, measured by the gradient of the Moreau envelope. Each round permits multiple simultaneous queries. In the stated parameter regime, the standard subgradient rate is optimal in sufficiently high dimensions, even for randomized algorithms making polynomially many parallel queries per round. We then show how the order structure of the real line enables faster search: a deterministic algorithm using only function values achieves logarithmic round complexity, with a further reduction through parallel queries. Matching lower bounds hold even when each query returns the entire subdifferential. Together, these results clarify the role of dimension in determining the benefits of parallel queries for weakly convex optimization.

fields
Optimization and Control
pdf ↗ page arXiv:2611.07675 ↗

Minimizing nonsmooth nonconvex objectives over general convex sets using no derivatives of the objective, together with structural access to the constraint set, poses a fundamental geometric difficulty: at boundary points, the feasible direction cone may fail to be closed, preventing the construction of finite sets of search directions that capture all feasible descent directions. We develop a feasible-directions framework for problems structured as the difference between a continuously differentiable function and a convex function over a closed convex set, and access the objective without knowledge of its decomposition, through exact line searches along feasible segments. To address the geometric obstruction, we introduce $η$-feasible directions, which remain uniformly separated from the boundary of the feasible direction cone, enabling finite-direction search even in the presence of nonlinear constraints. We present two algorithmic realizations of this framework. The first assumes access to an $η$-direction oracle and extends feasible-directions methods from linear to general convex constraints. The second replaces the oracle by a noise-augmented randomization scheme that constructs spanning directions from the affine-hull inner normal and almost surely avoids nonsmooth boundary points when initialized outside that set. Under curvature bounds and Lipschitz continuity of the smooth component's gradient, we jointly control the spanning coefficients and feasible step lengths of the oracle construction. This yields finite-time approximate $d$-stationarity along every unit feasible direction admitting a prescribed positive step length. For both methods, we also establish a rate in the proximal directional minimization (PDM) measure, assuming uniformly bounded direction-set sizes. Randomized exploration ensures that, almost surely, every accumulation point is $d$-stationary.

fields
Optimization and Control
pdf ↗ page arXiv:2611.07691 ↗

Heating, ventilation, and air-conditioning (HVAC) systems account for a significant portion of buildings' energy consumption, motivating the design of energy- and cost-efficient control strategies. However, conventional strategies employ simple, reactive control, without incorporating a model. Therefore, in this paper we adopt an economic model predictive control framework and focus on designing online algorithms to solve the control input design problems with a limited computational budget. In particular, we propose two tailored online algorithms based on projected gradient descent, with a custom projection routine, and the projected primal-dual method. We discuss the pros and cons of both algorithm designs, focusing in particular on computational efficiency. We further compare them through numerical experiments simulating the control of an HVAC system over the course of a day. The results show that computationally efficient online solvers for HVAC systems control are indeed possible. Moreover, the performance closely matches that of a baseline algorithm with unlimited computational budget, in terms of reference temperature tracking and comfort temperature range enforcement.

fields
Optimization and Control
pdf ↗ page arXiv:2611.07777 ↗

Linear Complementarity Problems (LCPs) with sufficient matrices form a well-studied class of problems, dating back from the foundational works of Cottle and Dantzig (1968). LCPs capture convex quadratic programs, and admit reductions from problems such as discounted stochastic games or mean-payoff games. While the original approach for solving these problems was based on pivoting methods, in their 1991 book, Kojima, Megiddo, Noma, and Yoshise introduced an interior point method (IPM) with running time that is polynomial in the input size, but which depends linearly on a certain parameter called the handicap number of the matrix. Despite the ensuing rich literature on sufficient LCP IPMs, the linear dependence on the handicap number has not been improved since then. We present an IPM whose running time depends on the square root of the handicap number. The algorithm can decrease the duality gap by a constant factor in $O\left(\min\left\{\sqrt{n} (1+κ),n^{9/8}\sqrt{1+κ}\right\}\right)$ iterations. Curiously, this can already be achieved using standard Newton updates. We analyze the IPM in a norm defined by a local weighting, where centralizing Newton steps quadratically converge to the central path, without any dependence on the handicap number. The weights arise from a characterization by Rump (Linear Algebra and Its Applications, 1997, 1998, 2003) of the sign-real spectral radius and its relation to sufficiency; another ingredient of our proof is a new operator norm characterization of the handicap number. Moreover, the connection to Rump's work also enables us to show coNP-completeness of testing sufficiency, and approximation hardness of the handicap number.

fields
Optimization and Control
pdf ↗ page arXiv:2611.07815 ↗

Accurate 4D reconstruction remains challenging: jointly estimating camera motion, scene geometry, and non-rigid deformation entails highly nonconvex optimization, making global optimality difficult to guarantee. We present XM4D, a framework for accurate and efficient 4D reconstruction built upon certifiably globally optimal geometric optimization. By separating the estimation of camera motion from non-rigid deformation, we formulate both stages as structured quadratically constrained quadratic programs (QCQP) that admit semidefinite programming (SDP) relaxations and verifiable global optimality certificates. For camera motion estimation, we lift static point tracks into 3D using foundational depth prediction models and formulate a scaled bundle adjustment problem that jointly estimates camera poses, depth scales, and static landmarks. By analytically eliminating translations and landmarks, we reduce the problem to a QCQP over scaled rotations. For non-rigid reconstruction, we represent dynamic motion using an embedded deformation graph (EDG) and show that its tracking, rigidity, and temporal regularization objectives similarly reduce to a rotation-only QCQP. Both problems admit SDP relaxations that we solve efficiently with a GPU-accelerated Burer-Monteiro solver based on Riemannian optimization. The recovered geometry further initializes dynamic Gaussian splatting for photometric refinement. Extensive experiments across seven benchmarks demonstrate accurate camera motion estimation on both static and dynamic scenes, efficient optimization that scales to thousands of frames, and state-of-the-art novel-view rendering quality.

fields
Optimization and Control
pdf ↗ page arXiv:2611.07902 ↗

We study how the spectral geometry of the mixing matrix enters the convergence of decentralized stochastic gradient tracking (GT) for smooth strongly convex problems. General bounds contain the factor $Δ_λ^{-3}$ in the higher-order stochastic term, where $Δ_λ$ is the absolute spectral gap. We show that, even for two agents with identical quadratic objectives, the exact steady-state mean-square disagreement scales as $α^2σ^2Δ_λ^{-3}$ as the nontrivial eigenvalue approaches $-1$, provided the stepsize $α$ is sufficiently small relative to the gap. Here $σ^2$ is the gradient-noise variance. By unrolling the disagreement recursion, we establish a stochastic bound that averages all nontrivial eigenvalue contributions for diagonalizable mixing matrices, up to an eigenvector conditioning factor. As a real or complex eigenvalue approaches the unit circle away from $1$, its contribution grows as the inverse cube of its distance to the circle; growth can be slower near $1$. For symmetric mixing with gaps $p$ at $1$ and $ν$ at $-1$, the weighted sum $θ_1/p+θ_2/ν^3$ replaces the prior factor $(Δ_λν^2)^{-1}$. For $N$ agents, $Nθ_1$ and $Nθ_2$ count the nonnegative and negative nontrivial eigenvalues, respectively. On a symmetric family, a finite-horizon lower bound for the average iterate matches the worst-case factor $p^{-1}+ν^{-3}$ up to universal constants. Numerical experiments confirm the predicted scaling.

fields
Optimization and Control
pdf ↗ page arXiv:2611.07958 ↗

Drivers are often told to back into a tight parking space, not drive in nose first. The advice looks paradoxical, because a slow car is dynamically reversible: reverse its speed and replay the steering backward, and it retraces a feasible path. Played backward, though, reverse-in shows a car driving forward out of the space. Reverse-in and forward-in are different tasks: they share a start position but end facing opposite ways. We separate three clearances between car and obstacles. Static fit $F$ is the clearance the parked car keeps at its best acceptable destination pose. Eventual access $A_\infty$ is the clearance it can keep all the way there with unlimited corrections. Resource-bounded access $A_K$ is the clearance that survives a budget of $K$ reversals, switches between forward and reverse. They satisfy $F\ge A_\infty\ge A_K$. In one fully specified scene (one car, bay, aisle and start), a certified argument proves $A_\infty=F=0.5000$ m for both tasks. In this scene, all certified forward/reverse asymmetry therefore arises at finite $K$: it lies in how many corrections each task needs, not in the clearance either can eventually keep. With one reversal, an explicit path gives reverse-in at least $0.4410$ m clearance. A certified reachability argument caps forward-in's one-reversal clearance at $d_0<0.1720002$ m. With two reversals, forward-in gets at least $0.2833$ m. To keep a clearance strictly above any $d$ with $d_0\le d<0.2833$ m, reverse-in thus needs exactly one reversal and forward-in exactly two when both tasks move forward first. If both move in reverse first, the counts are two and three. The contrast depends on the start position (Section 4.15). Proposed experiments test whether observers distinguish fit, eventual access and budgeted access. They compare these with simpler cues, such as how the turning body sweeps, and learned manoeuvre labels.

comments
79 pages, 15 figures
fields
Optimization and Control
pdf ↗ page arXiv:2611.07965 ↗

Benchmarking analogue optimisers requires certifying not only how close a returned energy is to the optimum, but also how many better configurations may remain. Semidefinite relaxations provide rigorous bounds on the ground-state energy and selected observables, yet reveal little about the size of this residual search space. We introduce a deterministic, instance-specific upper bound on the number of binary configurations whose energies improve upon a returned state by a prescribed margin. Combined with a semidefinite energy interval, this yields a complementary certificate of both the magnitude and multiplicity of any possible improvement. To judge an analogue optimiser's native performance, we provide a generalisation of the framework to obtain bounds on the Haar volume of the remaining continuous phase-space. To further characterise the remaining search space, we extend our bounds to resolve regions in configuration space in both the binary and continuous setting. Our approach couples a variational-Chernoff reduction with an elimination bound of the resulting partition function. Comparisons with exact enumeration and independent Haar sampling are consistent with the bounds on small instances. In numerical experiments at $N=1000$, even the largest family-median bound restricts the potentially improving binary configurations to less than $10^{-144}$ of the full configuration space, while the largest upper bound on the corresponding Haar volume is of order $10^{-499}$.

comments
17 pages, 5 figures
fields
Optimization and Control
pdf ↗ page arXiv:2611.07968 ↗

Port-Hamiltonian systems are an energy-based modelling formalism, that has strong ties to the physical models and allows interconnection of systems. Funnel MPC, a Model Predictive Control (MPC) scheme, allows for output tracking with prescribed error bounds. Earlier works focused on systems with stable internal dynamics, which typically excludes port-Hamiltonian systems. We introduce a modified stage cost design that takes into account the port-Hamiltonian structure of the system and allows deactivation of the funnel cost. We show initial and recursive feasibility for a class of port-Hamiltonian systems with relative degree one including unstable internal dynamics.

fields
Optimization and Control
pdf ↗ page arXiv:2611.08015 ↗

We propose and analyze a globalized stochastic semismooth Newton method for stochastic optimization problems in Hilbert spaces where the objective is strongly convex with Lipschitz continuous, Newton differentiable gradient. Gradient and generalized Hessian information are only available through mini-batching. In each iteration, a full semismooth Newton trial step is accepted or rejected based on a stochastic residual test and rejected steps are replaced by damped steps in the same Newton direction. Throughout, we admit gradient estimators with affinely growing variance, as arises in optimization problems with random partial differential equations as constraints. We prove non-asymptotic global R-linear convergence in expectation as well as almost sure convergence. For sufficiently fast-growing batch sizes, the method transitions almost surely into a regime of pure Newton steps and converges R-superlinearly almost surely. The method is applied to the optimal control of a random elliptic equation featuring a Huber tracking term as well as a random parabolic equation with a Moreau–Yosida penalization of an almost sure state constraint.

comments
25 pages, 3 figures
fields
Optimization and Control
pdf ↗ page arXiv:2611.08061 ↗

We present a model and an exact solution method for single-source capacitated facility location problems with decision-dependent demand uncertainty. The model allows the joint probability distribution of customer demands to depend on the facility to which demand points are allocated. The number of possible decision-dependent distributions is therefore equal to the number of feasible allocations of demand points to (groups of) facilities. To address the exponential size of the model, we apply and extend a recent version of the L-Shaped method for stochastic programs with decision-dependent uncertainty. In particular,we compare the original optimality cut with a new optimality cut based on a bilinear reformulation. We prove that, separated together with the original one, the new cut yields a tighter restriction, at the expense of introducing bilinear terms. In addition, we prove an effective decision-independent valid inequality that significantly improves convergence. Extensive numerical tests on 180 instances demonstrate that the method, supported by the valid inequality, can find optimal or near-optimal solutions to problem instances with up to 620 potential probability distributions. The tests also show that, while the new cut closes slightly more instances, the original cut remains preferable in terms of solution time and optimality gap.

fields
Optimization and Control
pdf ↗ page arXiv:2611.08090 ↗

Boolean networks (BNs) are widely used to model gene regulatory networks and other biological systems. In this paper, we investigate the minimal controllability problem for BNs subject to degree constraints. We first establish three new best-case upper bounds for XOR-BNs whose dependency graphs are $k$-regular. We then introduce classes of BNs that are invariant under cyclically partite transformation (CPT-invariant) and prove theorems connecting BN controllability with the surjectivity of vector-valued Boolean functions. Finally, by combining an analysis of this surjectivity with a probabilistic argument, we derive a new general lower bound on the minimum number of control nodes required for majority-type threshold BNs with $(2k+1)$-regular dependency graphs.

comments
16 pages
fields
Optimization and Control
pdf ↗ page arXiv:2611.08101 ↗

This work identifies a fundamental, controller-independent limitation in reusable launch vehicle powered descent: a sudden $180^\circ$ phase transition in the open-loop transfer functions from the control inputs to the lateral drift and drift rate. The transition occurs inside the flight envelope, reverses the effective control direction, and can destabilize the closed loop system under uncertainty. An analytical condition governing this transition is derived, its physical mechanism is established, and its dependence on vehicle mass, aerodynamic properties, and geometry is characterized. An Incremental Nonlinear Dynamic Inversion (INDI)-based control architecture is then developed to inherently address this limitation and is tuned via $\mathcal{H}_\infty$ Closed Loop Shaping to provide local robustness and performance guarantees. The internal dynamics introduced by the INDI inversion are analytically derived and proven to be stable throughout the flight envelope. The proposed controller is compared against a robust gain-scheduled baseline controller through linear stability analysis and an extensive nonlinear Monte Carlo campaign. Both controllers achieve 100% success across the considered Monte Carlo campaign, though the INDI-based controller provides improved tracking performance, superior nominal and dispersed stability margins, requires fewer controller design points, and reduces controller tuning complexity.

comments
28 pages, 16 figures, 5 tables. Accepted for presentation at the AIAA SciTech 2027 Forum, Orlando, FL, January 2027
fields
Optimization and Control
pdf ↗ page arXiv:2611.08121 ↗

Interactive decision support systems often require repeated preference judgments, although identifying the best alternative may not justify eliciting all \(n(n-1)/2\) pairwise comparisons. We propose ALEC (Adaptive Lexicographic Elicitation of Comparisons), which re-ranks the alternatives after each judgment and asks the first unobserved pair in the lexicographic order of the current ranking, using the Analytic Hierarchy Process as the concrete setting. Under a log-normal error model, directly comparing a contrast maximises the one-step reduction in its effective resistance. Posterior bounds on the probability of a wrong choice or of exceeding a relative-loss tolerance provide stopping rules valid under adaptive questioning, while a consistency-and-margin condition gives a deterministic certificate of agreement with the full-matrix winner. Simulations show that re-ranking is most useful in larger problems with limited budgets; ALEC outperforms Harker's original rule in most scenarios with \(n\ge11\), while under a common stopping rule it achieves higher hit-rate point estimates than the stronger Harker-log variant at a modest additional questioning cost. With a 5\% loss tolerance and a cap of \(2n\) additional comparisons, the stopping rule saves \res{prot-s06-saveadd-min}--\res{prot-s06-saveadd-max}\% of them, while among runs stopped by the bound the observed rate of losses above tolerance remains at most \res{s06-errearly-max}\%, against a nominal 10\%. ALEC therefore provides an adaptive elicitation module for decision support systems that balances decision precision against user effort.

fields
Optimization and Control
pdf ↗ page arXiv:2611.08123 ↗

In this paper, we study a penalty method for simple bilevel optimization in which the lower-level objective has a convex composite structure. Our approach is based on a vanishing-penalty reformulation, which gives rise to a sequence of penalized problems as the penalty parameter decreases. A key difficulty is that metric subregularity of the lower-level optimality mapping does not, in general, yield a uniform error bound for the penalized problems as the penalty parameter vanishes, preventing a direct application of standard proximal Newton theory. To address this difficulty, we first analyze a regularized proximal Newton step by exploiting the specific structure of simple bilevel problems, and establish the local estimates that underpin superlinear convergence. Building on this local analysis, we propose SiB-Newton, which incorporates the proximal Newton step into a globalization strategy. We establish convergence guarantees for SiB-Newton and show that, under suitable local conditions, Newton steps are eventually accepted and the iterates converge at a superlinear rate. Numerical experiments on synthetic examples verify the predicted local superlinear convergence rate.

fields
Optimization and Control
pdf ↗ page arXiv:2611.08297 ↗

Online Convex Optimization with memory and time-varying constraints (COCO-M) is a recently-introduced framework that captures the dynamics of stateful online learning and non-stochastic control with budget constraints. Despite its expressive power, COCO-M remains largely understudied. In this work, we tackle this problem in its full scale, providing an optimistic algorithm that incorporates untrusted predictions to tame universal dynamic regret and cumulative constraint violation. Our analysis leverages a potent reduction from memory to delay and treats state-dependent gradients as delayed feedback. We establish the first bounds that adapt to both environment non-stationarity (path length $P_T$) and prediction accuracy, while lifting previous restrictive assumptions such as functional separability. En route to these results, we derive findings of independent interest for the memory-to-delay reduction, and for delayed OCO with time-varying constraints. The proposed framework ensures sublinear worst-case guarantees that improve with prediction accuracy, reaching $\mathcal O(1)$ regret under perfect predictions, with cumulative constraint violation (CCV) $\mathcal O(\sqrt{T})$ for unknown $P_T$ and $\mathcal O(\log T)$ for known or online observable $P_T$ , while recovering the optimal COCO rates when memory disappears.

comments
Accepted at the 40th Conference on Neural Information Processing Systems (NeurIPS 2026)
fields
Optimization and Control
pdf ↗ page arXiv:2611.08302 ↗

We prove that integer linear feasibility is strongly $NP$-complete even when the constraint matrix has exactly two nonzero entries in every row and column, with its support forming a single cycle. This holds already for monotone systems (one positive and one negative coefficient in each inequality) with nonzero coefficients of absolute value $2$ or $3$ and right-hand sides in $\{-1,0,1,2\}$. No variable bounds are supplied separately from the system. For systems in which each inequality involves at most two variables, this strengthens known hardness results by restricting all numerical data to fixed finite sets. For systems in which each variable occurs in at most two inequalities, this establishes, to the best of our knowledge, the first $NP$-hardness result for feasibility. We also give negative answers, unless $P=NP$, to questions on polynomial-time solvability of integer programs with bounded coefficients and bounded totally $Δ$-modular treewidth, and on fixed-parameter tractability of integer bound propagation with respect to coefficient and right-hand-side magnitudes.

fields
Optimization and Control
pdf ↗ page arXiv:2611.08304 ↗

This paper analyzes a bilevel formulation for learning the regularization parameter of combined local-global (CLG) optical flow models. The lower-level problem computes the flow from a Lucas–Kanade data term and a quadratic or total variation (TV) regularizer, and the upper-level problem selects the parameter that minimizes the squared distance to a ground-truth flow. The setting is finite-dimensional, and a full-rank assumption makes the lower-level solution unique. For the quadratic model we derive sufficient conditions for the existence of an optimal parameter, for every optimal parameter to be positive, and for strict local optimality. For TV regularization an optimal parameter always exists, because the lower-level solution vanishes for large parameters, and we give a sufficient condition for positivity that accounts for the coupling of the two flow components through the local structure tensors. Synthetic examples illustrate the conditions, including cases in which they fail.

comments
31 pages, 5 figures. Submitted to the Journal of Mathematical Imaging and Vision. Python code in the ancillary files
fields
Optimization and Control
pdf ↗ page arXiv:2611.08359 ↗

We consider optimization problems with general equality and inequality constraints and partitioned cardinality constraints, which include complementarity constraints. In the augmented Lagrangian method that we consider, the bounds and the cardinality constraints are kept in the subproblems, which are solved by the Telescopic Method. The limit points of this method are B-stationary, without constraint qualifications, if the penalty parameters remain bounded, but not in general. The reason is that the swaps of the Telescopic Method pass through points at which the constraints have different values, and the corresponding variation of the penalty term grows with the penalty parameter. With unbounded penalty parameters and a derivative test outside the current face, the feasible limit points at which a linear independence constraint qualification holds are M-stationary for MPCCs. We modify the swaps, by means of a scaled Newton restoration, so that they preserve the values of the constraints, and we prove that every feasible limit point at which this constraint qualification holds is then B-stationary (S-stationary for MPCCs), whatever the behavior of the penalty parameters. Counterexamples show that the modification is necessary, even with exact solutions of the subproblems, and that the constraint qualification cannot be weakened to one of Mangasarian-Fromovitz type. In the MacMPEC collection, the method found certified S- or B-stationary points in 185 of the 193 instances, more than Algencan and the regularization method of Scholtes. In experiments with cardinality-constrained portfolios and sparse principal component analysis, the constraint-preserving swaps improved the original swaps and the methods applied to complementarity reformulations.

comments
41 pages, 2 figures, 6 tables. Supplementary material (results by instance) in the ancillary file bmalmpcc-suppl.pdf
fields
Optimization and Control
pdf ↗ page arXiv:2611.08404 ↗

Adversarial Training (AT) is an important method for improving model robustness. We study the dynamics of deterministic Gradient Descent (GD) in AT. Most existing works in this direction focus on the small or vanishing step sizes regime, where GD resembles a continuous-time flow. However, such step sizes can make training inefficient and are rarely used in practice. In this paper, we study the AT dynamics of GD with large constant step sizes. We first study GD in adversarial linear logistic regression, focusing on two aspects: where the model parameters converge (implicit bias) and how quickly the losses decrease (optimization). We then turn to neural networks training and study how the sharpness of losses evolve during AT. Our results advance the understanding of the behaviors of GD with large step sizes in AT and highlight three key differences from its behavior in Standard Training (ST): (i) the implicit bias gap between AT and ST, previously established only for small step sizes, persists for arbitrarily large constant step sizes, (ii) different from ST, there is no uniform rate of convergence for the optimization error in AT, and (iii) sharpness can exhibit two-stage dynamics that differ from the Edge-of-Stability behavior observed in ST.

fields
Optimization and Control
pdf ↗ page arXiv:2611.08457 ↗

We study composite optimization over the Bures–Wasserstein (BW) space of Gaussian measures. Our main observation is that a useful class of BW problems can be lifted into the Euclidean space of affine transport maps, such that the lifted problems can be solved efficiently using accelerated proximal gradient methods. For objectives whose smooth and nonsmooth components admit convex smooth and convex proximable affine lifts, respectively, our algorithm BW-FISTA achieves minimax-optimal objective gap $O(k^{-2})$ after $k$ iterations. We verify these regularity conditions for potential and interaction energies together with common spectral regularizers, and we treat stochastic gradients under a suitable noise model. As a principal application, we consider Gaussian variational inference with a log-concave, log-smooth target. Here, the resulting entropy term is handled by an efficient, exact proximal step, and we improve the previous rate of $O(k^{-1})$ to $O(k^{-2})$. We numerically verify this predicted boost with synthetic experiments.

comments
Submitted on September 2 to the Geometric Distributional Deep Learning workshop at NeurIPS 2026, accepted
fields
Optimization and Control
pdf ↗ page arXiv:2611.08468 ↗

Mathematical Physics34

In a previous work we conjectured that the multi-invariants of the ground state of a gapped $(d-1)+1$-dimensional system, labeled by bipartite $(d+1)$-edge-colored graphs encoding closed oriented $d$-manifolds $M$, compute the partition functions $Z(M)$ of the low energy topological quantum field theory (TQFT), up to non-universal factors determined by the local structure of the triangulation. In this paper we determine which combinations of partition functions can be isolated from the non-universal factors. The answer is governed by oriented bordism. If $M$ is null-bordant, $Z(M)$ is obtained from an explicit, state-independent finite family of multi-invariants, up to a power of $Z(S^d)$ which is fixed to be $Z(S^d)^{χ(M)/2}$ for even $d$ and can be removed for odd $d$. Conversely, the obstruction is measured by a quotient of the group of "residues" of colored graphs which receives a surjective map from the oriented bordism group, with kernel precisely the torsion subgroup as far as smooth manifolds are concerned. In $d=4$ the obstruction is the signature: $Z(M)/Z(S^4)^{χ(M)/2}$ can be extracted if and only if $σ(M)=0$. This is due to the fact that the first Pontryagin class $p_1$ has a local expression and thus $e^{iθσ(M)}=e^{\frac{i}{3}θ\int_M p_1}$ cannot be separated from the non-universal factors. Torsion invariants, in contrast, are not obstructed: for the ground state of the $4+1$-dimensional invertible phase with partition function $(-1)^{\int w_2w_3}$ we compute the multi-invariants exactly, and two graphs with identical local data, encoding the Dold manifold and $\mathbb{CP}^2\times S^1$, give opposite signs.

comments
65 pages, 14 figures
fields
Mathematical Physics, Algebraic Topology
pdf ↗ page arXiv:2611.07000 ↗

We develop a novel conformal bootstrap formulated directly in terms of conformal symmetry, reflection positivity, and crossing symmetry, without explicitly relying on the operator product expansion or conformal blocks. By constructing Gram matrices from states created by pairs of local operators and imposing spectral constraints through conformal generators, we obtain a hierarchy of semidefinite programs that directly constrain correlation functions. We demonstrate the method in one to three dimensions, obtaining correlator and spectral bounds that capture generalized free field behavior and reproduce the characteristic Ising kinks in two and three dimensions. We further show that joint positivity of states at different insertion configurations naturally constrains correlators away from self-dual kinematics.

comments
24 pages, 5 figures
fields
Mathematical Physics
pdf ↗ page arXiv:2611.07004 ↗

Coupling two-dimensional fundamental Dirac fermions to $SU(N)_k$ Chern-Simons theory produces non-chiral interfaces that in the infrared flow to a compact boson and a $W_N$ minimal model. We propose a description of these interfaces as continuum limits of $q$-deformed fermion chains. The $q$-deformation is achieved by coupling a chain of one-dimensional fundamental complex fermions to $SU(N)_k$ Chern-Simons theory, keeping the underlying topological gauge field continuous. The operator algebra is controlled by quantum groups, which provide an algebraic perspective on the usual categorical formulation of anyon chains. As an application, we derive a Lieb-Schultz-Mattis constraint involving baryon-number and lattice translation, obstructing a unique symmetric gapped ground state. At unit filling, we gap out the compact boson sector with a Mott term and obtain standard anyon chain Hamiltonians previously studied in the literature. This provides an anomaly argument for the existence of $N$ copies of minimal model sectors permuted by lattice translation in the infrared.

comments
65 pages + appendices
fields
Mathematical Physics
pdf ↗ page arXiv:2611.07012 ↗

We study how boundary correlation functions encode the charge dependence of the renormalised Euclidean action of an AdS$_3$ black brane. In three-dimensional Einstein-Maxwell theory, the Maxwell field has logarithmic asymptotics and is locally dual to a massless pseudoscalar. In the canonical ensemble, the charge density fixes the slope of a spatially linear source for the dual marginal operator. At fixed temperature, the classical action is even in the charge density due to reflection symmetry $Φ\to-Φ$, and its charge expansion coefficients are governed by Catalan numbers. We compute the quadratic and quartic terms from connected correlation functions around the neutral brane. The scalar two-point function and the shift symmetry of its source determine the matter Weyl anomaly, which fixes the logarithmic temperature dependence of the quadratic term in the charge expansion. At quartic order, the connected four-point function comes from tree-level graviton exchange. Integrating it against the linear source, with an infrared cutoff on the overall scale of the integral, reproduces the quartic coefficient of the action per unit length. At each higher order, the corresponding Catalan coefficient predicts the integrated sum of all connected tree-level Witten diagrams.

comments
41 pages, 4 appendices
fields
Mathematical Physics
pdf ↗ page arXiv:2611.07019 ↗

Structural-response surrogates must capture moving resonances, phase changes, and unequal-amplitude transfer channels during optimization or health-state changes. Physics-embedded learning seeks FEA-level accuracy at lower cost by embedding physics in inputs, forward passes, losses, and backpropagation. We propose Modal Rational (MR), mapping structural features to modal quantities for analytical frequency-response reconstruction. The Physics-Guided Modal Operator Network (PGMON) adds bounded frequency, damping, and residual-flexibility corrections. Modal Operator with Structural GELU and Anchored-SIREN Correction (MOSAIC) denotes the bounded spectral-discrepancy framework used here; MOSAIC-MR and MOSAIC-PGMON add such discrepancies. They support finite-dimensional maps and physical decoders for future neural operators; evidence is fixed-topology, not a completed operator comparison. Two 20,000-design populations are studied: an 80 m steel truss tower with variable member areas and attached masses, and a tapered shell beam with ten variable station masses. Archived tests cover ten designs and seven acceleration histories per structure; they are benchmarks, not seismic-design assessments. With 5% reference modal damping, MR, PGMON, and MOSAIC give mean displacement-history errors of 0.73-0.76% for the tower and 0.029-0.116% for the beam, against declared band-limited modal references. Supplied modal frequencies are distinguished from design-only use; unequal parameter counts and training settings are explicit. Error analysis separates learned modal discrepancies, truncation, and excitation weighting, explaining acceleration and difficult-tower limits and risks of blind use. Results support physically structured prediction but limit claims about generalization, computational savings, and added corrections. Next is developing and validating neural operators around these architectures.

fields
Mathematical Physics
pdf ↗ page arXiv:2611.07059 ↗

We consider a dual version of heterotic Kodaira-Spencer theory on a Calabi-Yau three-fold $X$ with background fluxes, related to the type I BCOV theory of Costello-Li. We implement the Green-Schwarz mechanism of Costello-Li, where the Beltrami differential plays the role of the Green-Schwarz field. On a compact Calabi-Yau three-fold with background gauge fluxes $F$, where $\mathrm{tr}(F^2)$ is trivial in cohomology, retaining gauge anomalies only and leaving gravitational anomalies to future work, we incorporate the Costello-Li Green-Schwarz mechanism into the classical deformation complex. This gives a theory governed by a nilpotent operator $\bar D$ on $Q=T^{1,0}X\oplus\operatorname{End}(V)\oplus T^{*1,0}X$, whose extension maps carry holomorphic derivatives, so that $(Q,\bar D)$ is not a holomorphic bundle. Using diagonal $\bar\partial^\dagger$-gauge, the one-loop gauge anomaly cancels, assuming that the extension maps of $\bar D$ do not modify it, which we argue is plausible. Introducing an auxiliary $(0,2)$-form, we construct a cubic interacting theory whose gauge symmetries are governed by an $L_3$ algebra. On compact Calabi-Yau three-folds, and for sufficiently small fields, we show that the equation of motion is equivalent to the Maurer-Cartan equation, together with a divergence-free constraint on the Beltrami differential. Finally we trivialise $\bar D$ locally to the Dolbeault operator. The (non-tensorial) transition maps glue into a quasi-holomorphic extension sheaf $\widetilde{\mathscr{Q}}$, for which we prove a Dolbeault theorem, $H^{0,q}_{\bar D}(Q)\cong\check H^q(\widetilde{\mathscr{Q}})$.

comments
59 pages, 2 appendices
fields
Mathematical Physics, Differential Geometry
pdf ↗ page arXiv:2611.07178 ↗

In the note, we use the principles of non-symmetric integral geometry (non-integral geometry), related to the non-symmetric restricted integration measure, to extend the well-known relations traced from the Fourier slice theorem and the dual Radon transform. The proposed generalization is a very useful for the alternative restoration of the outset function. The latter impacts entirely on the image reconstruction process. The derived relations confirm the complexity of the inversion operator thanks for the symmetry breaking of integration measure.

comments
9 pages in JHEP style
fields
Mathematical Physics, Functional Analysis
pdf ↗ page arXiv:2611.07376 ↗

A recent remarkable paper by OpenAI showed that the spin-one Heisenberg chain with periodic boundary conditions has a spectral gap in the thermodynamic limit. The proof elegantly deduces this gap from the Hamiltonian's spectrum on a finite system. This basic idea is an inequality involving the purity of the original Gibbs state and a dual Gibbs state that is defined via space-time rotation. Here we consider a few generalizations of this purity bootstrap argument. We first extend it to arbitrary spatial dimensions. Our main result uses the space-time duality between symmetry generators and symmetry twists to obtain upper bounds on the splitting of nearly degenerate low-energy states, together with a nonzero gap above them, from finite-system information. As applications, we show that finite-system calculations for the 2+1-D toric code in a field and the 2+1-D transverse-field Ising model establish parameter regimes with four and two low-lying states whose splittings are bounded from above by $e^{-cL}$ and $e^{-cL^2}$ respectively for arbitrarily large system size $L$, while the gap above these states remains nonzero in the thermodynamic limit. Next, we extend it to fermionic systems and use a finite calculation to establish a symmetric mass generation phase in a $2+1$-D fermionic model. Finally, we apply the bootstrap to tensor-network states and establish exponential decay of connected correlations of all local operators in the symmetric square-lattice spin-one state introduced by Jian and Zaletel.

comments
15 pages main text + Appendices
fields
Mathematical Physics
pdf ↗ page arXiv:2611.07400 ↗

Symmetry governs the organization of physical structures, while symmetry breaking reveals their transformations. In geometric deep learning, symmetry imposes rank constraints on equivariant representations, limiting the expressive capacity of equivariant graph neural networks (GNNs). Here we show that these constraints can instead be exploited: the singular-value spectra of equivariant representations provide a unified fingerprint of structural symmetry. We introduce spectral symmetry fingerprints (SymFP), which enable discrete point-group identification through rank-template matching and continuous symmetry analysis by quantifying spectral energy beyond target rank constraints. SymFP is orientation invariant and architecture agnostic, enabling a parameter-free GNN implementation. Its spectra can be reused across multiple target groups, and the resulting leakage scores satisfy subgroup monotonicity. Across synthetic structures, molecules, coordination complexes and large biological assemblies, SymFP accurately identifies point groups, tracks controlled symmetry breaking and reveals local and degree-resolved symmetry changes that can be obscured by global geometric measures. These results establish the spectral structure of equivariant representations as a general framework for characterizing symmetry across physical systems.

fields
Mathematical Physics
pdf ↗ page arXiv:2611.07410 ↗

Quantum resources can change the dynamical features and operational behaviors of conventional networks. Long-distance entanglement supports information transfer between nodes lacking a direct connection. This entanglement link not only modifies the stationary effective connectivity of the network but can also be used to suppress undesirable classical dynamics. By examining information flow control mechanism, we demonstrate that classical collective behavior can be well-balanced by altering the initially distributed entanglement. We show the entanglement thresholds for two distinct control approaches: linear response control and Lyapunov functional controllability.

fields
Mathematical Physics
pdf ↗ page arXiv:2611.07440 ↗

Following the knot-quiver correspondence, we investigate quiver counterparts of the HOMFLYPT skein relations by interpreting link invariants in terms of holomorphic worldsheets and their multicover contributions. Extending this framework to more general worldsheet geometries, we conjecture that, up to quiver equivalence, the adjacency matrix of the quiver associated with any link in a skein triple admits a block decomposition whose diagonal blocks are quiver adjacency matrices associated with the other two links. We verify this conjecture in numerous examples.

comments
46 pages, 5 figures; a Mathematica file is attached as an ancillary file
fields
Mathematical Physics, Geometric Topology, Quantum Algebra, Symplectic Geometry
pdf ↗ page arXiv:2611.07455 ↗

We apply the logarithm construction of higher-derivative conformal primary (super) fields proposed by Butter et al. to derive simple Wess-Zumino actions for: (i) Weyl anomalies in arbitrary even dimensions; (ii) $\mathcal{N} =1$ and $\mathcal{N}=2$ super-Weyl anomalies in four dimensions; and (iii) $(p,q)$ super-Weyl anomalies in two dimensions, with $p,q \leq 2$.

comments
22 pages
fields
Mathematical Physics
pdf ↗ page arXiv:2611.07547 ↗

We investigate the collisional quantum dynamics of an open system in contact with a thermal bath. We propose a particle-bath interaction model based on the Weyl quantization principle and the Wigner phase space description of quantum evolution. By factorizing the reduced Hamiltonian of the open system via supersymmetric quantum mechanics techniques, we construct system-dependent creation and annihilation operators for arbitrary potential landscapes. This allows us to extend the classical Fokker–Planck relaxation mechanism to incorporate Lindblad quantum stochastic jumps within the Wigner phase-space formalism. Through a semiclassical expansion of the Moyal star product to second order, we obtain a generalized Wigner–Lindblad kinetic model featuring position-dependent friction and diffusion terms. We establish the mathematical well-posedness of the Lindblad problem and we validate our approach through numerical simulations. Our model resolves unphysical long-time artifacts of standard quantum Fokker–Planck models in quartic double-well traps, accurately captures dissipative quantum state localization in 2D molecular potentials (4D phase space), and effectively evaluates electron transport and trapping times in InAs/InP resonant tunneling diodes.

fields
Mathematical Physics
pdf ↗ page arXiv:2611.07608 ↗

We establish conditions for exactness of the inverse renormalization group via constrained Markov processes, a class of Monte Carlo simulations which sample the degrees of freedom within a block on a fine lattice under the constraint that they remain consistent with the corresponding degree of freedom on the coarse lattice according to a forward renormalization group transformation. We show analytically, for an arbitrary transformation, that constrained Markov processes correspond to an exact inversion of the renormalization group at the fixed point action, and to a near-exact inversion in the vicinity of the fixed point. We then optimize renormalization group transformations directly on inversely renormalized configurations to effectively map the action onto itself and ensure computational stability in the generation of configurations for high-dimensional systems. Starting from a set of uncorrelated configurations for the three-dimensional $φ^{4}$ quantum field theory at lattice volume $V=16^{3}$, we invert the renormalization group with constrained Monte Carlo to construct a set of uncorrelated, near-equilibrium configurations for $V=128^{3}$ and extract two critical exponents of the system. The inversion exhibits no critical slowing down, with dynamic critical exponent $z=0.03(2)$.

fields
Mathematical Physics
pdf ↗ page arXiv:2611.07641 ↗

Several closed-form non-axisymmetric magnetic fields with nested toroidal flux surfaces have recently been discovered as exact benchmarks for verifying three-dimensional magnetohydrodynamic equilibrium codes. However, all of these solutions have at most two field periods ($N \le 2$), i.e., identical segments that repeat toroidally. Here, we derive closed-form non-axisymmetric magnetohydrodynamic equilibria for an arbitrary integer number of field periods $N \geq 1$. The equilibria have non-constant pressure and closed magnetic field lines with either zero rotational transform ($ι= 0$) for $N \geq 1$ or integer rotational transform ($ι= \pm N$) for $N \ge 3$. For the reflection-symmetric $ι= 0$ equilibria, we also construct a matched exterior vacuum field with continuous magnetic field across the plasma boundary, providing a benchmark for free-boundary equilibrium codes.

fields
Mathematical Physics
pdf ↗ page arXiv:2611.07672 ↗

A composite boson made of two fermions can look elementary under one measurement and composite under another. In an exactly solvable paired-mode model we compare three probes: classical drives, periodically binned quadrature statistics, and intensity correlations $g^{(2)}$. The Weyl defect has an exact product formula and a controlled dilute expansion in the Schmidt moments. The binned observables are instead piecewise constant in the Schmidt weights and sensitive to single eigenvalues crossing bin edges; on the flat-spectrum ladder they converge slowly and nonmonotonically over the computed range. A spectral filter and a polar correction recover an exact Pauli algebra; a heralded joint projection then extracts a logical singlet from the vacua of two independent copies of the composite, with probability set by the accepted weight and the mixedness of the filtered reduced qubit. Finite-mode examples with Schmidt number about three succeed at the percent level despite order-one Weyl defects; near-unit $g^{(2)}$ does not control the binned algebra. For flat spectra the construction converges strongly to its oscillator counterpart, whose vacuum overlap is strictly positive at every scale; for infinite spectra with summable Schmidt amplitudes it converges at all but countably many filter cutoffs, positivity at the cutoffs studied remaining open.

comments
27 pages, 5 figures, 7 tables
fields
Mathematical Physics
pdf ↗ page arXiv:2611.07776 ↗

We describe a symplectic reduction of the Zhukovskiy integrable case of motion of a rigid body about a fixed point with a gyroscope in $\mathbb R^n$ to an integrable magnetic flow on a sphere $S^{n-1}$. We also present Zhukovskiy's geometric interpretation of motion of the Euler top with a gyroscope, aka the Zhukovskiy–Volterra gyroscope, and perform a symplectic reduction of this system to an integrable magnetic flow on a $2$-dimensional sphere.

comments
17 pages, 4 figures, the material in Section 2 appeared in the first version of arXiv:2601.03965
fields
Mathematical Physics, Differential Geometry
pdf ↗ page arXiv:2611.07789 ↗

Non-local couplings offer a means of engineering the dynamics and topology of discrete lattices beyond the constraints of nearest-neighbour (NN) interactions. Here, we extend the mass-spring Su-Schrieffer-Heeger (SSH) model to incorporate beyond-nearest-neighbour (BNN) connections that preserve the chiral symmetry of the diatomic unit cell, investigating how their range and strength shape the bulk response. We show that the accessible winding numbers are bounded by the lattice connectivity, while the spring stiffnesses select the specific topological phase. We further investigate the challenges posed by non-local interfaces, where the extended connectivity inherently introduces local boundary perturbations. The dynamic freedom afforded by BNN couplings is exploited by examining interfaces between dynamically and topologically dissimilar bulk media, revealing a band-gap-induced asymmetry in the localisation of interface states. We show that the interfacial configuration provides a means of controlling the existence and character of localised interface modes, with changes to the interface enabling the tuning of mode frequencies, localisation and hybridisation with the bulk bands.

comments
14 pages, 8 figures
fields
Mathematical Physics
pdf ↗ page arXiv:2611.07885 ↗

Topological recursion is a universal procedure that generates an infinite family of multi-differentials with different interpretations in various areas of mathematics. Recently, a new understanding of topological recursion has emerged from $x$-$y$ duality, which relates two such families through explicit formulae. We study $x$-$y$ duality, its parametrised version, and symplectic duality in the setting of equivariant topological recursion. We explicitly analyse the pole structure of the resulting dual families and provide recursive formulae for the different types of polar parts. In contrast to classical topological recursion, the dual family has additional poles at fixed points of the group action, and at anti-diagonals determined by the group action. We show that the partition function $Z_P(E)=\int_{H_N} e^{-N\,\mathrm{Tr}(EΦ^2+P(Φ))}dΦ$ of the Kontsevich matrix model with arbitrary potential $P$, where $dΦ$ is the Lebesgue measure on Hermitian $N\times N$ matrices and $E$ a positive matrix, admits a Fock space realisation related to the character expansion of the IZHC integral. The Fock space framework provides two mutually $X$-$Y$ dual families of multi-differentials relative to an implicitly defined spectral curve $(X,Y)$. The solution $Y(X)$ succeeds for the quartic Kontsevich model with potential $P(Φ)=\fracλ{4}Φ^4$. We identify a network of duality transformations and symplectic shifts which relate the quartic Kontsevich model with the $\mathbb{Z}/2\mathbb{Z}$-equivariant spectral curve. The proposed construction yields a conjectural all-genus solution to the quartic Kontsevich model, which reproduces recursive structures previously established for genus $g\leq 1$.

comments
95 pages (appendix 40 pages)
fields
Mathematical Physics, Algebraic Geometry, Combinatorics
pdf ↗ page arXiv:2611.08018 ↗

Ostracism was the practice in use in Athens and other ancient Greek cities to send into exile citizens whose influence was deemed to be dangerous. Quite similarly, splay ostracism is here the approximation that bans splay distortion from thin films of ferroelectric nematic liquid crystals on a curved shell in space. This reduces to a mere geometric problem the quest for equilibria of director vortexes and the walls isolating them, as splay ostracism requires the director field lines to be everywhere orthogonal to systems of geodesics on the shell. Consequences of this theorem are explored for shells with Gaussian curvature of either sign. The resulting quasi static dynamics reveals both cases where walls escape from the shell and cases where they are trapped on it, with possibly different reflections on the vortex dynamics.

fields
Mathematical Physics, Differential Geometry
pdf ↗ page arXiv:2611.08027 ↗

The square-lattice Ising model is exactly solvable in zero magnetic field, but the scope of exact solutions at nonzero field remains unresolved. We show that, at every fixed dimensionless field $H=h/(k_BT)\ne 0$, its free energy and magnetization obey no algebraic differential equation in the coupling $K=J/(k_BT)$. This excludes Onsager-type formulas and broad classes of special-function expressions. Previous work left open a countable set of fields with algebraic-integer fugacities. We eliminate these exceptions using Galois conjugation and the large-order growth of droplets of reversed spins. At conjugate fugacities outside the unit circle, nearly square droplets produce coefficient growth incompatible with an algebraic differential equation. We prove that cancellations cannot remove this growth in the full pressure series. The free-energy result also holds for the Lieb lattice and square lattices with equal-length chain decorations and uniform coupling and field. In every dimension $d\ge2$, we additionally resolve fields for which $e^{-2|H|}$ has a real algebraic conjugate greater than one.

comments
17 pages, 2 figures
fields
Mathematical Physics
pdf ↗ page arXiv:2611.08033 ↗

Tumour evolution and treatment resistance are driven by the interplay between clonal heterogeneity, selective pressures, and therapeutic interventions. In this work, we propose an integrated mathematical and statistical framework to investigate clonal selection, predict radioresistance, and optimize radiotherapy protocols. We develop a multi-clonal dynamical model incorporating proliferation, resource competition, epigenetic instability, and clone-specific radiosensitivity. Through the statistical analysis of 255,150 in silico simulations, we identify key biological determinants of treatment response and characterize the effects of evolutionary competition and radiotherapy on clonal composition. Our results show that evolutionary dynamics favour highly proliferative clones, whereas radiotherapy selects radioresistant populations and reduces tumour heterogeneity. We further investigate how tumour characteristics influence resistance trajectories and identify trade-offs between tumour control, toxicity, and resistance selection across different treatment schedules. Optimal protocols are determined for distinct patient profiles, accounting for dose, fractionation, and timing. Finally, we demonstrate how treatment history affects the efficacy of second-line therapies. Overall, our findings highlight the potential of combining mechanistic modelling and statistical inference to understand cancer evolution and support personalized radiotherapy design.

fields
Mathematical Physics
pdf ↗ page arXiv:2611.08059 ↗

We prove exponentially slow thermalization for the quantum random energy model (QREM) under Davies dynamics. The QREM is the $n$-qubit Hamiltonian $H=ΓT+U$, where $U$ assigns independent Gaussian energies with variance equal to the system size to computational-basis configurations and $T$ is the transverse field generated by single-qubit Pauli $X$ operators. Let $β_c=\sqrt{2\log 2}$ be the freezing inverse temperature of the classical random energy model. For every inverse temperature $β>β_c$ and every transverse-field strength $0\leqΓ<Γ_c(β)$, where $e^{ββ_c}=2\cosh(βΓ_c(β))$, we show that the mixing time of the dissipative Davies dynamics associated with single-qubit Pauli system–bath couplings grows exponentially with the system size, almost surely; the bound is unchanged if the standard zero-Bohr-frequency dephasing terms are included. The result is uniform over all Davies rates satisfying detailed balance and a polynomial upper bound and does not assume primitivity. In the spin-glass phase, the Gibbs weight is carried by a sparse family of deep classical configurations that are separated by an extensive Hamming distance. A localization theorem for the QREM [Manai, Warzel, Comm. Math. Phys. 402, 1259 (2023)] transfers this geometry to the corresponding low-energy quantum eigenstates. Consequently, the Gibbs weight of the intermediate spectral sector is exponentially small, while single-qubit bath operators have exponentially small matrix elements between the two macroscopic sides of the bottleneck. A Davies-specific flux estimate then converts these two facts into an exponential lower bound on the mixing time. This gives a rigorous open-system dynamical manifestation of the frozen, localized low-energy structure of the QREM throughout its equilibrium spin-glass phase below the quantum transition line.

fields
Mathematical Physics
pdf ↗ page arXiv:2611.08064 ↗

Thermodynamic curvature is widely used as a diagnostic of interactions and phase transitions, yet its interpretation is usually tested only in the thermodynamic limit. Small spin clusters are physically relevant in their own right, since a macroscopic sample of independent molecular trimers or a few-qubit quantum simulator is governed exactly by the few-spin partition function. They also provide the only setting in which the geometry can be obtained in closed form together with its full dependence on the interaction topology. We study the Fisher-Rao (Ruppeiner) geometry of three Ising spins with arbitrary pairwise couplings in a longitudinal field. This cluster interpolates continuously between an open chain and a geometrically frustrated triangle. For the symmetric triangle we obtain the scalar curvature in closed form. It vanishes on the noninteracting line and on a line of fixed field-to-temperature ratio that does not depend on the coupling, which coincides with the inflection point of the single-spin susceptibility. The near-zero behaviour is therefore exactly linear and reflects analyticity rather than critical scaling. In the frustrated antiferromagnetic regime the curvature remains finite and negative down to zero temperature. A general low-temperature formula relates this plateau to the change in magnetization content between the degenerate ground manifold and its first excitation. Any bond anisotropy destroys the plateau below a crossover temperature that depends only logarithmically on the anisotropy, so the curvature is an extremely sensitive detector of exact ground-state degeneracy. On the ferromagnetic side the zero-temperature curvature is universal for all couplings, whereas additional zero lines in anisotropic clusters track field-induced crossings of excited levels.

comments
17 pages, 5 figures
fields
Mathematical Physics
pdf ↗ page arXiv:2611.08072 ↗

We prove multifractal scaling of measures related to the eigenfunctions of Schrödinger operators with point-like bump potentials, in the thermodynamic limit, in dimension $d\geq 5$. This is consistent with a longstanding expectation in the physics literature in these dimensions. The fractal exponents are computed explicitly as a function of the parameter defining the moments. Our result requires a weak disorder condition on the location of the scatterers and a strong coupling condition on the sequence of eigenvalues.

comments
25 pages, no figures
fields
Mathematical Physics
pdf ↗ page arXiv:2611.08087 ↗

We develop a Pohlmeyer-type reduction for classical non-relativistic strings on string Newton-Cartan backgrounds. For strings on a two-sphere, we identify the invariant reduced field, show that its dynamics is integrable, and construct a Lax representation. We relate this intrinsic formulation to the non-relativistic expansion of relativistic strings and obtain a dictionary between non-relativistic string solutions and their reduced counterparts. The construction explicitly determines which relativistic solitonic configurations survive the non-relativistic limit and which decouple. Moreover, it provides a framework for applying integrability methods to Newton-Cartan string dynamics.

comments
22 pages. Waiting for comments!
fields
Mathematical Physics
pdf ↗ page arXiv:2611.08164 ↗

Overcompressive shocks in a $2\times2$ hyperbolic system of conservation laws describing small-amplitude longitudinal-torsional waves in an elastic rod are investigated. A viscous regularization method with an anisotropic dissipative matrix is applied to the system. A qualitative analysis of phase portraits has revealed that a two-parameter family of heteroclinic trajectories, representing the structure of an overcompressive shock, exists in one of six domains in the parameter plane. It is established that the model admits overcompressive shocks, which belong to the first type according to the classification of Liu.

fields
Mathematical Physics
pdf ↗ page arXiv:2611.08292 ↗

We present a systematic framework for gauging continuous symmetries, and show that the Schur-Weyl duality arises naturally as a realization of this construction with open boundary conditions. Restricting to compact Lie group symmetries, we construct dual models for one-dimensional quantum lattice models exhibiting those symmetries. We demonstrate that matrix product operators enable the explicit construction of intertwiners between the dual theories and their (non-invertible) symmetries. With closed boundary conditions, we can relate the full spectrum of these dual theories by considering all possible twisted boundary conditions.We illustrate how the Peter-Weyl transform appears when the duality acts on the boundary Hilbert space of a $G$-symmetric model. The presented framework generalizes previous work on dualities for unitary fusion category symmetric models. It exemplifies how dualities for both finite and continuous symmetries rely on the relation between a bond algebra and a representation of the symmetry in the lattice as double centralizers. We argue that dual non-invertible spontaneous symmetry breaking is the mechanism that underlies symmetric tensor network algorithms. We illustrate these results for $U(1)$- and $SU(2)$-symmetric quantum lattice models.

fields
Mathematical Physics
pdf ↗ page arXiv:2611.08293 ↗

We consider a prototypical class of dynamics of $N\gg 1$ degrees of freedom subject to two competing terms, one favoring relaxation to a trivial equilibrium and one taking the form of a Gaussian force field promoting the emergence of many local equilibria. The latter is a non-gradient field, i.e. it includes a non-reciprocal component with a tunable strength. For this model, we compute the quenched complexity, i.e. the exponential growth rate of the typical number of equilibria, resolved by stability index. We compare it with the annealed complexity, i.e. the exponential growth rate of the average number of equilibria. We show that the two coincide only in the fully non-reciprocal limit: for any finite degree of reciprocity, typical realizations contain exponentially fewer equilibria than predicted by the average count. Increasing reciprocity enhances the discrepancy between the two. The quenched complexity depends explicitly on the covariance of the random force field and on the degree of non-reciprocity: the universality observed at the annealed level breaks down. The average counting of equilibria correctly identifies the location of the topology trivialization transition, but it does not capture the typical organization of equilibria near this transition, nor the stability index of the dominant equilibria.

comments
25 pages (6 main + 19 SI), 5 figures (3+2)
fields
Mathematical Physics
pdf ↗ page arXiv:2611.08324 ↗

We introduce the notion of an oriented $\vee$-system which gives a class of solutions of the oriented associativity equations generalising Veselov's $\vee$-system solutions of WDVV equations. An oriented $\vee$-system is a special configuration of pairs of a vector and a covector in a linear space. We show that the class of oriented $\vee$-systems is closed under taking subsystems and restrictions. We provide examples which include a multiparameter family related to imprimitive complex reflection groups.

comments
24 pages
fields
Mathematical Physics, Differential Geometry
pdf ↗ page arXiv:2611.08326 ↗

We study the long-time, large-distance asymptotic behaviour of the two-point correlation function of fields of the impenetrable Bose gas in equilibrium states characterized by a quasi-momentum distribution (or filling fraction). This includes, in particular, the cases of Gibbs and generalized Gibbs ensembles. The overall structure of the asymptotic expansion depends on the number of Fermi points, those points for which the quasi-momentum distribution takes the value $1/2$. The leading asymptotic behaviour is determined by a relaxation rate, a logarithmic correction term and an amplitude for which we provide simple and explicit expressions. In the case of two Fermi points the first sub-leading correction is calculated as well.

comments
43 pages, 16 figures. Submission to SciPost Physics
fields
Mathematical Physics
pdf ↗ page arXiv:2611.08350 ↗

A signal band-limited to one rectangular band and observed through a finite window is the classical concentration problem of Slepian, Landau, Pollak, and Widom. The eigenvalues of the operator cluster near one and near zero, separated by a transition called the plunge. The plunge width grows like the logarithm of the window length, with a universal coefficient of one over pi squared set by the band-edge jump. This paper follows the clusters and the plunge as the band is deformed. An elementwise power of the kernel smooths the band edge, and the plunge gives way to a power-law edge with exponent one less than the power. Disjoint copies of the band widen the plunge in proportion to the number of band edges, as the multiple-interval theorem of Landau and Widom implies, so the coefficient is a per-edge constant. Shrinking each band to a line gives the classical finite-rank operator. Overlapping bands give a multilevel staircase, and a self-similar band gives a cluster hierarchy governed by the Cantor dimension. At the smooth edge, rescaled eigenvalues do not collapse onto one curve. The edge descriptor passes from a constant to an exponent to an integer to a fractal dimension. For the single band, classical results place all eigenfilter zeros on the unit circle, with more of them inside the band as the order rises, so the dominant eigenvectors act as lowpass filters.

comments
26 pages, 10 figures
fields
Mathematical Physics, Classical Analysis and ODEs, Spectral Theory
pdf ↗ page arXiv:2611.08401 ↗

Sampling experiments are a leading approach to demonstrating quantum computational advantage, yet their robustness to experimental noise is not completely understood. We develop a common framework for classical simulation of boson and fermion sampling under thermal loss, including pure loss. For Fock-state boson sampling, Gaussian boson sampling with fixed uniform squeezing, and fermion sampling with magic inputs, we obtain polynomial-time sampling at inverse-polynomial total variation error, for any fixed nonzero overall loss rate, and with high probability over Haar-random interferometers. A milder flatness assumption on the ideal output distribution suffices for quasipolynomial-time simulation without Haar randomness. With constant thermal loss per layer, the same flatness assumption gives polynomial-time simulation at logarithmic depth. Our bounds for Fock-state boson sampling cover both dilute and saturated regimes and become stronger near full input filling: when only a bounded number of input modes are empty, the noisy output for Haar-random interferometers converges, as the number of modes grows, to a distribution independent of the interferometer.

comments
83 pages, 2 figures
fields
Mathematical Physics
pdf ↗ page arXiv:2611.08444 ↗

We provide an explicit scheme to realize a quantum memory in any 2D non-chiral topological order. For this, we construct syndrome extraction circuits based on the Turaev-Viro path integral as well as a classical algorithm for decoding on the fly. We prove that our scheme has a threshold under a local stochastic circuit-level noise model and discuss how to generalize it to implement topological logical gates.

comments
37 pages, 6 figures
fields
Mathematical Physics
pdf ↗ page arXiv:2611.08488 ↗

Information Theory27

Large-scale signal processing systems, including aeronautical and aerospace platforms, increasingly rely on tensor operations, whose distributed execution is constrained by communication, memory, and latency bottlenecks. In this paper, we propose a tensor-aware coded computation framework for distributed mode-1 tensor–matrix multiplication that operates directly on tensor subtensors, avoiding explicit unfolding and preserving multi-dimensional structure. For a fixed partitioning configuration, the proposed tensor-PolyDot scheme achieves the same recovery threshold, communication cost, worker-side computation, and memory requirements as conventional matrix-based PolyDot schemes. However, by leveraging multivariate encoding, it replaces high-degree univariate interpolation at the fusion node with structured low-dimensional decoding. This leads to a substantial reduction in decoding complexity and latency, without affecting any other system metric. Additionally, the tensor-based approach improves memory locality and enables parallel decoding across tensor modes, making it well-suited to modern hardware architectures. Numerical results show decoding speedups proportional to the tensor partitioning factor, reaching up to an order-of-magnitude improvement in practical settings. These advantages make the proposed framework particularly relevant to latency-critical aeronautical and aerospace distributed signal processing systems.

comments
To appear in Proceedings, IEEE International Workshop on Signal Processing Systems (SiPS), November 2nd-4th, 2026, Bordeaux, France
fields
Information Theory
pdf ↗ page arXiv:2609.30567 ↗

We propose a joint framework for constructing and decoding Logically Augmented (LA) quantum low-density parity-check (QLDPC) codes. This is motivated by the need for flexible coding rates and adaptive error protection while preserving block length and code structure. On the construction side, we use an auxiliary \textit{Logical Protection Code (LPC)} to impose additional stabilizer constraints on selected logical qubits of a \textit{primary} code. On the decoding side, the proposed decoder uses the LPC syndrome to generate logical error hypotheses and condition belief propagation (BP) on each hypothesis. The proposed \textit{Hierarchical Logical List Decoder (HLLD)} is invoked only when the initial BP decoder for primary code fails to converge. It then selects the correction with the smallest Hamming weight among the syndrome-consistent candidates produced by hypothesis-conditioned BP runs. For numerical evaluation, we focus on short repetition LPCs, which provide a flexible means of trading rate for additional error protection. Each LPC syndrome admits exactly two logical error hypotheses, irrespective of repetition length, which enables HLLD to explore both possibilities without relying on a single logical estimate. Numerical results show that varying the repetition length provides a tunable trade-off between rate and error-correction performance across a wide range of primary code constructions and their operating regimes. On the decoding side, with the overall LA code held fixed, HLLD reduces the logical error rate by up to 81\%, 75\%, and 69\% for LA codes based on mitten, lifted-product, and bivariate bicycle codes, respectively, compared with conventional BP decoding of the same LA codes.

fields
Information Theory
pdf ↗ page arXiv:2611.07062 ↗

Large language models (LLMs) are increasingly recognized for their remarkable capabilities; federated learning (FL) offers a promising framework for collaboratively fine-tuning LLMs on edge devices while achieving data privacy. However, edge devices are inherently heterogeneous: clients may exhibit diverse data distributions and resource constraints, including limited memory and computation budgets. Recent federated Mixture-of-Experts (MoE) approaches address these challenges by using sparse expert activation or partial expert loading. However, existing partial-expert methods rely on heuristic full-pool observations or prioritize system-wide load balance, without guarantees for recovering client-specific relevant experts from partial observations. We propose $S^3$-MoE, a server-side sparse expert assignment framework that estimates client-specific expert relevance from partial observations and refines assignments through an exploit–explore mechanism. We also establish relevance-estimation error and exact Top-$K$ recovery guarantees. Experiments using a CNN-based MoE on CIFAR-100 show that $S^3$-MoE recovers task-dependent client–expert preferences and reduces conflicts caused by correlated expert selection across clients. Compared with Full-20 FedAvg, which gives each client access to all 20 experts for local Top-5 routing, $S^3$-MoE achieves 1.31 percentage points higher mean final accuracy while loading only $5$ experts per client.

comments
Accepted for publication in the proceedings of the 2026 IEEE Information Theory Workshop (ITW 2026)
fields
Information Theory
pdf ↗ page arXiv:2611.07093 ↗

We introduce dithered belief-propagation decoding for circuit-level quantum memory to mitigate the oscillations and trapping behavior that limit standard belief-propagation (BP) decoders for qLDPC codes. Our approach combines a quantum-adapted sign-flip mechanism with chained randomized prior perturbation. Evaluated under circuit-level noise on the distance-11 rotated surface code and the distance-12 bivariate-bicycle code, the proposed dithered decoder significantly outperforms BP and BP+OSD+CS-10. The proposed decoder achieves accuracy comparable to or better than state-of-the-art Relay-BP, especially at higher physical error rates, while trading a moderate increase in additions for about 98% fewer multiplications on average and eliminating node-level RNGs. These results suggest that dithering is a promising approach toward real-time decoding of quantum memories.

comments
7 pages, 5 figures. In the proceedings of QCE 2026
fields
Information Theory
pdf ↗ page arXiv:2611.07144 ↗

We study the age of information and peak age of information in status update systems with an arbitrary number of independent sources sharing a single server with generally distributed service times. We consider a general age-dependent self-preemptive policy in which an in-service update from source $c$ is preempted by a new same-source arrival with a probability $p_c(x)$ that depends on the elapsed service time $x$. This formulation subsumes the non-preemptive, fully self-preemptive, and probabilistically preemptive policies as special cases, and also accommodates threshold-based preemption. We develop a unified analytical framework based on non-homogeneous absorbing continuous-time Markov chains and semi-Markov chains to derive explicit expressions for the average age of information and peak age of information of each source under general per-source service-time distributions and age-dependent preemption rules. The framework applies to both single-source and multi-source systems and enables joint tuning of per-source preemption parameters. Numerical results for fifteen service-time distributions in the single-source setting reveal that the benefits of preemption depend strongly on the service-time hazard-rate structure, with threshold-based preemption providing its most notable additional gains over age-independent policies for distributions with unimodal hazard rates. For a dual-source system with lognormal service times, jointly tuned threshold-based preemption achieves lower sum average age of information and peak age of information than the considered age-independent policies, highlighting the importance of cross-source interactions in preemption decisions.

fields
Information Theory
pdf ↗ page arXiv:2611.07151 ↗

We consider lossless source coding with a marginal distribution constraint on the encoding symbols (SC-MDC). Unlike distribution matching, signal shaping and homophonic coding which require the symbols to be (approximately) i.i.d. following a target distribution, SC-MDC requires each symbol to have a marginal distribution that exactly matches the target distribution, which can be regarded as a perfect perceptual quality constraint on the symbols, and may have applications to wear leveling, source coding with cost and steganography. We give nonasymptotic lower and upper bounds on the optimal encoding length, and exactly characterize the coding rate (or the first-order term of the length) when the source distribution or the target distribution is uniform, or when the source is an i.i.d. sequence.

comments
15 pages
fields
Information Theory
pdf ↗ page arXiv:2611.07156 ↗

Power matters for AI safety. Optimal agents tend to acquire power across a wide range of objectives (instrumental convergence). As we embed AI agents into socio-technical systems, humans risk gradual disempowerment. Both concerns call for a formal, quantitative measure of power. Two such measures in the context of Markov games are information-theoretic empowerment, the capacity of the channel from an agent's actions to future states, and capability-based individual power, a measure of goal attainment built from the probabilities that a human would attain each goal in a family of candidates were they pursuing it. We show where the two meet. From individual power, Fano's inequality yields a floor $B(s)$. This floor is a lower bound on what we term goal revelation, the mutual information $\mathbb{I}(G;ξ\mid s)$ between the human's latent goal and the realised trajectory. Empowerment caps revelation from above, so that $B(s) \le \mathbb{I}(G;ξ\mid s) \le \mathbb{I}(\mathbf{A}_h \to ξ\mid s) \le \mathfrak{E}_h(s)$. Each quantity can be large while the one below it stays at zero, highlighting the following failure modes. An assistant maximising empowerment can prefer an interface offering broad influence over goal-irrelevant degrees of freedom to one that lets the human attain every goal they might hold. An assistant maximising revelation cannot tell a working interface from one rewired so that every goal yields the wrong outcome, since relabelling goals leaves mutual information unchanged while zeroing every attainment probability. Positive guarantees therefore come only from the attainment end of the chain. A satisfaction floor, computable from two statistics of interaction logs, offers a way to monitor the power users retain in a deployed system, and simulations in random environments show the three quantities are correlated but far from interchangeable.

comments
19 pages, 3 tables. Code: https://github.com/RuizSerra/revealing-goals
fields
Information Theory
pdf ↗ page arXiv:2611.07247 ↗

In vehicular integrated sensing and communication (ISAC) networks, each service-provider vehicle uses a limited power budget to serve multiple sensing targets and communication receivers. Maximizing the communication sum-rate subject to sensing signal-to-interference-plus-noise ratio (SINR) constraints is nonconvex, and iterative optimization is costly to rerun for every network layout. This paper proposes a graph neural network (GNN)-based power allocation algorithm, termed GNN-ISAC. Concretely, GNN-ISAC addresses the aforementioned problem by considering two distinct edge types in the graph representation, one capturing the inter-provider interference, the other representing the intra-provider power-budget sharing. GNN-ISAC is trained without labels using a sensing-penalized sum-rate loss. We also adapt the weighted mean-square error minimization (WMMSE) algorithm to this problem by introducing a sensing-power reservation step. Simulation results show that GNN-ISAC achieves communication sum-rates close to WMMSE with low sensing-link violation rates and much less computation. It also outperforms an existing heterogeneous GNN and a channel-blind baseline, and generalizes to larger networks without retraining.

comments
Submitted to IEEE ICC 2027
fields
Information Theory
pdf ↗ page arXiv:2611.07277 ↗

We consider private retrieval of the matrix product $A^\top B^{(θ)}$ from replicated $P_c$-colluding servers when $B^{(\ell)}=A+N^{(\ell)}$, with the innovations mutually independent and independent of the user-held $A\in\mathbb{F}_q^{m\times m}$. The servers store the full library, correlated with $A$, which changes both the communication problem and the security definition: a secrecy condition including the stored library cannot hold, as the library itself partly reveals $A$. We therefore measure only additional leakage and require index privacy conditioned on $A$; both reduce to classical notions under independence. For $A$ of rank $k<m$ we prove, with no structural hypothesis on the map, $D^*\geq H(A^\top B^{(θ)}\mid A)/C_{\mathrm{PIR}}(N,P_c,L)-o(km)$, where $C_{\mathrm{PIR}}(N,P_c,L)$ is the replicated $P_c$-colluding private-information-retrieval capacity. The proof rests on a colluding converse for servers storing arbitrary refinements of the demanded messages, whose answers need not be measurable functions of the functional-message library, and a privacy-transfer characterization. Publishing a canonical row-space representative makes the bound achievable for arbitrary independent and identically distributed innovations under a public-setup measure. Under a strengthened observational model (exact correctness, common conditional library support, and full-transcript security), concealment forces $D\geq H(G_m N^{(θ)})/C_{\mathrm{PIR}}$, where $G_m$ spans the linear join of all row spaces in the support of the operand distribution; this extends to vanishing error under a simultaneous-decodability assumption. In the full-rank case product recovery reduces to innovation recovery: $D^*=H(N^{(θ)})/C_{\mathrm{PIR}}+o(m^2)$.

comments
36 pages, 4 tables
fields
Information Theory
pdf ↗ page arXiv:2611.07325 ↗

Within the Thompson VLSI model, the first energy optimal decoding schemes (up to polylogarithmic factors) that also reach clock cycle lower bounds are constructed. Let $N$ be block length. Then, there is an exponential error scaling function and a binary symmetric channel, such that for all $f(N)$ that scales to zero more slowly than this, there is a decoding scheme for this channel with error probability that scales as $O(f(N))$ that has number of clock cycles $T\le O(\sqrt{-\log(f(N)})\text{polylog}(N))$ and energy $E\le O(N\sqrt{-\log(f(N))}\text{polylog}(N))$. This matches fundamental lower bounds up to polylogarithmic factors. The technique uses expander-graph based LDPC codes, a mesh network architecture, and a parallelization technique. The key technical result needed for this is a lemma in graph theory. This lemma is used to show that communication graphs of bounded node degree can be routed through two- and three-dimensional mesh networks without overflowing the memory of any mesh node. Upper bounds are also extended to three dimensions.

fields
Information Theory
pdf ↗ page arXiv:2611.07336 ↗

Speech communication plays a crucial role in the Internet of Things and autonomous networks. However, traditional communication systems suffer from insufficient semantic extraction and limited noise suppression capability. To address these issues, we propose a Speech enhancement-based Deep Semantic Autoencoder (S-DSA) communication system for wireless communications. It consists of a dual-branch semantic encoder, a channel encoder, a channel decoder, and two semantic decoders, enabling joint source-channel coding for robust transmission. A key novelty lies in an adaptive semantic noise decoupling mechanism that disentangles target speech semantics from noise representations through an interaction module, thus maintaining semantic consistency under low SNR conditions. Furthermore, a joint training strategy optimizes speech reconstruction and noise estimation, while a lightweight testing strategy reduces complexity by disabling the noise-decoding path. The proposed S-DSA enables more effective separation between speech and noise, and yields higher robustness as well as better reconstruction quality. These findings highlight the potential of the proposed S-DSA as a practical solution for next-generation intelligent wireless communications.

journal
IEEE Transactions on Vehicular Technology, vol. 75, no. 7, pp. 13450-13462, 2026
fields
Information Theory
pdf ↗ page arXiv:2611.07361 ↗

For classical-quantum multiple access channels (cq-MACs) with two senders, we give several general expressions for the capacity region, valid for arbitrary sequences of channels. In the case of asymptotically vanishing error ($\varepsilon=0$), the capacity region is expressed in terms of spectral inf-mutual information rates. For every $0\le\varepsilon<1$, the $\varepsilon$-capacity region is expressed in terms of a quantity defined through a linear combination of three alternative hypotheses. In particular, the outer bound of Kubo and Nagaoka (2015) is tight. Our results include the quantum counterpart of the general formula of Han (1998) for the classical MAC. The most important ingredient is the one-shot inner bound of Sen (2021), which we apply to each member of an arbitrary channel sequence.

comments
11 pages
fields
Information Theory
pdf ↗ page arXiv:2611.07462 ↗

Reordering low-Earth-orbit (LEO) beam-hopping visits can reduce grant-free update delivery even when each region receives the same service time. We derive the largest limit on inter-frame position changes that meets a prescribed delivery-rate loss tolerance when every group is visited once per frame. The guarantee concerns a protected beam's spatially averaged long-run delivery rate relative to optimized periodic service under exogenous re-planning. Fixed admission can favor irregular visits, whereas admission optimized for each revisit interval makes the per-visit reward concave under sufficient conditions on transmission success. Periodic service is then optimal at a fixed mean interval. The shortest and longest allowed intervals determine the exact worst-case loss, attained by feasible alternating visit orders that respect every group's displacement limit. The resulting budget requires no estimate of the revisit-interval distribution and extends to independent stationary renewal updates. In a collision example with 40 groups and 40 resources per slot, increasing the loss tolerance from zero to 1% raises the limit from 17 to 24 positions. With independent priorities, the latter retains over 90% of the priority-weighted waiting-cost reduction achieved by unrestricted ordering. Received-power simulations under Rayleigh fading support the predicted loss.

comments
6 pages, 5 figures, 1 table, submitted to IEEE ICC 2027
fields
Information Theory
pdf ↗ page arXiv:2611.07592 ↗

For infinitely many rates $R \in (0,1)$, we show that full-length Reed-Solomon codes over fields of low characteristic can contain super-polynomially many codewords in a Hamming ball of radius just exceeding the Johnson radius. Over prime fields, we exhibit full-length Reed-Solomon codes of constant rate $R$ with polynomially many codewords in a Hamming ball of relative radius bounded strictly below $1-R$. This refutes a recent conjecture that, over prime fields, list sizes below capacity are always bounded by a constant. Together, these results demonstrate that certain restrictions in recent works on explicit Reed-Solomon codes achieving list-decoding capacity, related to the choice of field or evaluation points or to the list-size guarantee, are inherent.

comments
24 pages
fields
Information Theory
pdf ↗ page arXiv:2611.07607 ↗

The proliferation of computation-intensive applications, such as generative artificial intelligence and low-altitude economy services, demands mobile edge computing (MEC) systems that are both energy-efficient and adaptable to highly dynamic workloads. However, conventional MEC architectures rely on fixed-position antennas with limited spatial flexibility and rigid single-slot task execution models, which fail to accommodate bursty traffic and time-varying channels. To address these challenges, this paper proposes a novel MEC framework empowered by movable antennas (MAs) and task buffering. Unlike traditional approaches, our system enables continuous antenna repositioning to maximize array gains and decouples task generation from immediate execution via buffer queues, allowing for flexible multi-slot scheduling. We formulate a stochastic optimization problem to minimize long-term energy consumption by jointly optimizing receive beamforming, MA positions, and computing resources. Leveraging Lyapunov optimization, we transform the long-term problem into tractable per-slot subproblems and develop an efficient alternating optimization algorithm. Simulation results demonstrate that the proposed scheme significantly outperforms the benchmark schemes in terms of energy consumption, while maintaining stable task queues under unpredictable traffic conditions.

fields
Information Theory
pdf ↗ page arXiv:2611.07656 ↗

Thresholds for local properties tell us which guarantees to expect from random linear codes (RLCs) and help us carry those guarantees over to structured codes. A monotone property is $b$-local if a collection of at most $b$ codewords can witness its satisfaction or failure, as in list decoding and list recovery. The threshold of $\mathcal P$ is the code dimension at which the probability that an RLC in $\mathbb{F}_q^n$ satisfies $\mathcal P$ crosses $1/2$ in a sharp transition. We show that the threshold of a $b$-local property $\mathcal P$ is within $O(b)$ dimensions of the expectation threshold. This is the dimension predicted by the best union bound over covers of $\mathcal P$: families of subspaces such that the span of every witness contains at least one member. This gives a general derandomization statement: Any decreasing local property satisfied by RLCs at a given rate holds at a similar rate for pseudorandom ensembles that fool the cover union bound. In addition to previously studied ensembles, these include codes whose parity-check matrix is sampled from a small-bias distribution. Because our threshold characterization applies to all local properties, this derandomization also reaches properties beyond the scope of earlier reductions, including maximum loads in linear hashing for arbitrary input sets, rank-metric list decoding, and list decoding from insertion-deletion errors and burst errors. Our theorem unifies and strengthens prior works on row-symmetric and coordinate-wise linear characterizations, which located the threshold to within $O(q^{2b}\log_q n)$ and $O(b^2+nb/\log_2 q)$ dimensions, respectively. We obtain $O(b)$ error for every $b$-local property, independently of alphabet size. Under suitable pseudorandomness bounds, the rate loss vanishes for every $b=o(n)$, compared to $b=O(\log n)$ for earlier frameworks.

fields
Information Theory, Combinatorics
pdf ↗ page arXiv:2611.07661 ↗

Decisions about dense populations depend on how much will cross a boundary by a deadline and when the first member arrives. Conventional radar forms a range-Doppler map, whose cells add noise and whose detections merge unresolved members. We introduce over-the-scene computation, where illumination makes the scene compute such functionals without resolving members. With random echo phases, a correlator's mean output energy sums member reflectivities weighted by a kernel set by the waveform and template. The waveform's ambiguity function is this computation's transfer function, determining computable functionals and their noise cost. One readout aggregates members over a small delay-Doppler area but along a line segment growing with duration, longest for the chirp of matching slope among constant-envelope waveforms. Members predicted to share a range lie on such a line, so a chirp's mean beat spectrum is the predicted range profile at a look-ahead horizon set by its rate. For unbiased estimation, a Cramer-Rao bound over all equal-energy waveforms favors resolving members above unit signal-to-noise ratio (SNR) and grouping weaker ones into unit-SNR sets. At 0.65 members per cell, -10 dB member SNR, and one sweep, the chirp's bin sum inside the boundary comes within 16% of this bound in root-mean-square error and, with 77 times fewer samples per horizon, outperforms every unbiased estimator from a coherent processing interval of equal dwell, bandwidth, and energy. No finite set of waveform-template probes determines an unconstrained Rayleigh-fluctuating population's first arrival, but bisection localizes it if first arrivals outweigh later arrivals' sidelobes.

fields
Information Theory
pdf ↗ page arXiv:2611.07732 ↗

Dephasing is a prominent source of decoherence in quantum communication, degrading phase coherence and transmission reliability. Therefore, knowledge of the physical channel parameters for reliable transmission is imperative, motivating the integration of quantum communication and sensing at the physical layer. Consequently, we study the transmission of coherent states over a von Mises dephasing channel affected by thermal noise. We consider a heterodyne receiver, for which, we first derive the symbol-error probability (SEP) and an accurate closed-form approximation, through which an irreducible asymptotic error floor due to the dephasing channel is identified. We note that the derived SEP is affected by the mismatch between the receiver's mean phase estimation and the true mean phase, thereby formalizing the interplay between sensing accuracy and communication reliability. In addition, we derive the Fisher information (FI) for the mean phase acquisition of the channel and provide closed-form lower and upper FI bounds. Moreover, the quantum FI (QFI) of the channel is studied, and lower and upper bounds on the QFI are given, which become asymptotically tight as the photon number increases. Despite the initial complexity of the FI and the QFI, the bounds given are relatively simple and effective. In addition, by comparing the FI with the QFI, we show that heterodyne detection approaches the quantum sensing limit as the number of photons increase, while an unavoidable sensing floor exists due to dephasing. Then, using the SEP and the FI analysis, we discuss the conditions under which communication and phase estimation requirements are jointly feasible. Finally, numerical results validate the derived SEP, estimation bounds, and the sensing-communication feasibility tradeoff.

fields
Information Theory
pdf ↗ page arXiv:2611.07855 ↗

In goal-oriented communications, indirect source coding involves two sources: a semantic source $X$ and an observation $Z$, which is the only source available at the encoder. In this paper, we consider this framework for classification, where the decoder only needs to reconstruct $X$. We argue that the gap between the indirect rate-distortion function and the performance of practical learned coding schemes, such as variational information bottleneck (VIB) and variational autoencoder (VAE) architectures, comes from three effects: the latent representation, the choice of the deep neural network (DNN) architecture and training loss, and whether the encoder is aware of the task. To quantify these effects, we introduce four operational rate-distortion functions that account for the DNN architecture and the training loss, and we establish a hierarchy among them. We then derive these functions for a Gaussian mixture source, where $X$ is the mixture state. Numerical results, both for the Gaussian mixture and for practical coding schemes on the CIFAR-100 dataset, quantify the rate loss incurred by each effect.

comments
8 pages, 4 figures
fields
Information Theory
pdf ↗ page arXiv:2611.07946 ↗

Reliable intra-swarm communication is fundamental to cooperative sensing, control, and mission execution in unmanned aerial vehicle (UAV) swarms. In open low-altitude shared-spectrum environments, coexisting UAVs may cause benign interference through spectrum reuse and signal leakage, while jamming UAVs may deliberately degrade critical intra-swarm links. Since a jamming UAV can adjust its transmit power to disguise signal-to-interference-plus-noise ratio (SINR) degradation as link fluctuation in benign interference, stealthy jammer UAV detection remains a challenge. In this paper, we propose a novel belief-driven stealthy jammer UAV detection scheme from the perspective of Bayesian security game. To capture stealthy interference in shared-spectrum UAV swarms, the interfering UAV is modeled as a hidden benign or jamming type, whose transmit power strategy induces the SINR observed by the defender. The defender updates its posterior belief from the observed SINR and decides whether to maintain normal operation or trigger defensive power boosting. We characterize pure-strategy perfect Bayesian equilibrium (PBE) conditions under separating structures with distinguishable type and pooling structures with indistinguishable observations. Numerical results show that the proposed scheme improves the defender's expected utility compared to the conventional schemes.

comments
Accepted for presentation at IEEE GLOBECOM 2026
fields
Information Theory
pdf ↗ page arXiv:2611.07977 ↗

Partial information decomposition (PID) splits the information that two sources carry about a target into redundant, unique and synergistic parts, with nonnegative decompositions lying on a segment where redundancy measures are bounded by the positive part of the co-information and the smaller source-target information. We parameterize this segment by the overlap, the common excess of redundancy and synergy over their net balance, and show that for many of the named PID measures, their redundancy can be written as the largest positive part of the co-information over a set of reference distributions. This common form induces an ordering of those PIDs on the segment. With the smallest redundancy and synergy and the largest unique atoms of any nonnegative PID, the information-theoretic version of Hi-Fi, a feature-importance method, is the zero-overlap end of the segment for two sources, with the maximum overlap bound given by the minimum mutual information (MMI) PID. We introduce, by splitting the co-information into a pairwise maximum- entropy part and the third-order connected information, a PID that sits at zero overlap except when pairwise redundancy and triadic synergy coexist. For three or more sources the bivariate results hold on every pair of merged source groups, and among all nonnegative solutions of the lattice equations MMI and Hi-Fi still give the smallest and largest unique atoms but zero overlap on all pairs at once is often unattainable. Two inequalities across pairs are necessary in general and sufficient for three sources. For identity targets zero overlap is possible exactly when the sources' information diagram is nonnegative. In random three- and four-source distributions zero overlap on all pairs fails in 21-100% of the cases.

fields
Information Theory
pdf ↗ page arXiv:2611.08131 ↗

Predictive interference management supports ultra-reliable low-latency communication (URLLC) by allocating finite-blocklength resources against a high conditional quantile of next-slot interference. Building on an extreme-value-theory/kernel-density-estimation (EVT/KDE) predictor with quantile-defined interference states, this paper makes the predictor resilient to unannounced shifts in the interference distribution. Its cumulative refitting on all past observations does not specify how the state partition and retained samples should evolve after a regime change. We show that drift first makes the quantile partition inefficient: cumulative refitting preserves tail support and coverage but cannot recover conditional resolution, and pays for it in channel use, whereas discarding history starves the tail estimator. We therefore propose event-triggered quantile-space adaptation. A nonparametric cumulative sum (CUSUM) on the predictor's level stream detects change. At each confirmed event, the quantile partition is re-derived from post-change observations, retained history is re-binned under it, and each state is refit from a bounded per-state memory, capped to balance stale-regime bias against tail-estimation variance, with a closed-form L-moment tail estimator. At a nominal coverage target $η\!=\!0.99$, relative to matched cumulative refitting with fixed states, the proposed method reduces channel use by $5.7$--$13.2\%$ and quantile pinball loss by $25$--$50\%$ across abrupt, incremental, compound, downward, and reversal drift, with coverage changing by $-0.5$ to $+1.0$ percentage points. At matched achieved coverage, it saves $4$--$18\%$ of channel uses, while resource use in load-only and stationary cases changes by less than $0.5\%$. With $K$ samples per state and $L$ states, predictor memory is bounded at $O(KL)$ rather than growing as $O(t)$.

comments
Submitted to IEEE Transactions on Wireless Communications
fields
Information Theory
pdf ↗ page arXiv:2611.08132 ↗

Semantic communication (SemCom) using a relay node is considered, where the relay assists in conveying task-relevant information from a source to a destination, with meaning defined by classification labels. A joint end-to-end design of the source encoder, relay transformation, and destination decoders is developed to optimize both semantic inference accuracy and perceptual reconstruction quality for image data. Relay-assisted schemes are shown to outperform direct transmission, with gains that become more pronounced in low signal-to-noise ratio (SNR) regimes across different relaying strategies. Alongside these performance benefits, however, a key security vulnerability introduced by the relay is identified, and a Byzantine attack is formulated in which the relay injects structured, learned perturbations into its transmitted signal. The attack operates on intermediate communication representations subject to channel effects, with perturbations generated through a learned mapping applied at test time. It is shown to significantly degrade semantic inference accuracy at the destination. These results reveal a clear trade-off between SemCom gains and security risks in relay-assisted systems, motivating the need for secure relay designs that account for insider threats.

fields
Information Theory
pdf ↗ page arXiv:2611.08155 ↗

A collection of classical codes $C_1,\dots,C_m$ has product expansion if every low-weight word in the sum of their extensions to the product space can be written as a sum of words supported on few lines, each line being a codeword of the corresponding $C_i$. For $m=2$ product expansion is equivalent to robust testability of the tensor code, while for $m\ge 3$ it is strictly stronger; it is the hypothesis on the local codes behind recent constructions of good quantum LDPC codes, good quantum locally testable codes, and quantum codes with transversal non-Clifford gates. No explicit family over a fixed field was previously known to achieve constant rate, linear distance for both the code and its dual, and 3-fold product expansion at growing length. We prove that for the one-point evaluation codes $C_i$ of length $N_i=728\cdot27^{\,i-1}$ and pole order $\lfloor N_i/5\rfloor$ on the Garcia–Stichtenoth tower over $\mathbb{F}_{729}$, the triple $(C_i,C_i,C_i)$ has product expansion with a constant independent of $i$. The dual codes $C_i^\perp$ then form an explicit fixed-field family of rate tending to $109/130$ and relative distance at least $8/65$. The proof is geometric: we view the words of the sum code as arrays on the product of three copies of the $i$-th curve of the tower, where the parity checks are sections of a line bundle. An exact generation theorem describes the sum-code words supported on the grid footprint of a suitable divisor. A support-localisation argument then shows that every low-weight array lies on such a divisor, after deleting a small number of coordinates. We also give a sufficient criterion for three-fold product expansion in terms of degree and genus ratios, yielding explicit families over every non-prime finite field of order at least $512$. The principal conclusions have been formally verified in Lean 4.

comments
76 pages
fields
Information Theory, Algebraic Geometry
pdf ↗ page arXiv:2611.08172 ↗

This paper investigates a non-orthogonal multiple access assisted coordinated beam hopping (NOMA-CBH) scheme in multidimensional very low Earth orbit (VLEO) satellite communication systems, where multiple satellites employ CBH and coordinated beams jointly serve ground users and a dedicated ground station (GS). To characterize the VLEO-specific propagation environment, we develop a channel model that incorporates atmospheric attenuation, Doppler effects, and pointing perturbations. Building upon this architecture and channel model, a low-complexity decoupled (LCD) algorithm is designed to optimize beam scheduling and power allocation. We further derive closed-form outage probability expressions for the ground users and the GS, perform asymptotic analysis to characterize their diversity orders, and evaluate the delay-limited throughput. Simulation results demonstrate that: i) The NOMA-CBH with LCD algorithm achieves an enhanced system sum rate relative to orthogonal multiple access with coordinated beam hopping (OMA-CBH); ii) Across the considered heterogeneous traffic-demand operating points, the NOMA-CBH scheme achieves higher aggregate delivered traffic than the benchmark schemes; and iii) The outage performance of NOMA-CBH is better than that of OMA-CBH.

comments
16 pages. Published in IEEE Transactions on Aerospace and Electronic Systems (Early Access, 2026)
fields
Information Theory
pdf ↗ page arXiv:2611.08174 ↗

We determine the maximum number of common zeros in projective space of any set of linearly independent homogeneous polynomials of the same degree, when the degree is smaller than the cardinality of the finite field. This proves the conjecture of Beelen, Datta and Ghorpade and determines the complete weight hierarchy of the corresponding projective Reed-Muller codes. The proof uses Green's hyperplane restriction bound, combined with a footprint argument that reduces the problem to the affine case solved by Heijnen and Pellikaan. Wei duality then determines the hierarchy in sufficiently large degrees, recovering the complete hierarchy in every degree for the projective plane. Finally, we prove an upper bound valid in every degree, and conjecture that it is always sharp.

fields
Information Theory, Commutative Algebra, Algebraic Geometry
pdf ↗ page arXiv:2611.08203 ↗

The simplification of complex dynamical systems is essential for efficient estimation and control. A principled approach to this task is through the lens of information-theoretic compression. The Past-Future Information Bottleneck (PF-IB) provides a theoretical framework for compressing linear dynamical systems. However, the standard formulation often lacks structural consistency, yielding representations that are not realizable as dynamical systems. By constraining the objective to a specific manifold of block Hankel matrices, which we prove is a Lie group, we bridge the gap between PF-IB and realization theory. Our approach ensures that the resulting compressed dynamics are realizable, thereby enabling its direct application to estimation and control. We demonstrate our approach in numerical simulations.

fields
Information Theory
pdf ↗ page arXiv:2611.08426 ↗

Differential Geometry26

In this paper, we introduce the coupled free energy $F_{-γ}^{\rm cp}$ which is a natural generalization of the free energy functional defined by Berman. In the Fano case, our coupled free energy provides a diagonal approach to the coercivity of the coupled Ding functional. We also define the non-Archimedean coupled free energy $F_{-γ}^{\rm cp,NA}$ on the space of finite energy measures on the Berkovich analytification. As a consequence of the regularization process of the non-Archimedean entropy established by Boucksom-Jonsson, Chi Li and Trusiani, we prove that the existence of a unique coupled Kähler-Einstein metric is equivalent to the stability condition that there exists $ε>0$ such that $F_{-1-ε}^{\rm cp,NA} \geq 0$ on the space of measures associated with a test configuration for the anticanonical line bundle. As a coupled version of the work by Berman, we show that the law of the empirical measure satisfies a large deviation principle whose rate function is the coupled free energy under the assumption that a conjecture by Hultgren is true. We also introduce a coupled version of the microscopic stability thresholds and prove that a uniformly coupled Gibbs stable Fano manifold has a unique coupled Kähler-Einstein metric. This is a generalization of the results by Fujita-Odaka and Berman.

comments
57 pages. Comments are welcome
fields
Differential Geometry
pdf ↗ page arXiv:2611.06968 ↗

In the Möbius geometry of submanifolds there exists an important Möbius invariant tensor called the Möbius metric which together with the Möbius second fundamental form, the Blaschke tensor and the Möbius form constitute a complete system of conformal invariants for umbilic-free submanifolds when the dimension $n \geq 3$. An umbilic-free immersion $f:M^{n}\to \mathbb{R}^{m}$ is said to have constant Möbius Ricci curvature if its Möbius metric has constant Ricci curvature. In this paper, we classify umbilic-free immersions $f:M^{n}\to \mathbb{R}^{m}$ with constant Möbius Ricci curvature, flat normal bundle and two distinct principal normal vector fields. We prove that a submanifold with constant Möbius Ricci curvature and flat normal bundle has closed Möbius form, and then we classify such submanifolds provided the Möbius form vanishes identically on $M^{n}$.

comments
17 pages
fields
Differential Geometry
pdf ↗ page arXiv:2611.07152 ↗

We prove that a complete noncompact gradient shrinking Ricci soliton of dimension at least two is a finite quotient of an Einstein cylinder if its scalar curvature is bounded below by the cylindrical value. The corresponding pointwise strict bound implies compactness. We also establish one-endedness criteria using curvature information at critical points of the potential, including degenerate ones, and obtain sharp one-endedness and rigidity results under global Ricci norm bounds.

fields
Differential Geometry
pdf ↗ page arXiv:2611.07381 ↗

We study holomorphic sectional curvature and the Calabi curvature operator on compact Kähler manifolds. First we show that every homogeneous Kähler metric on a generalized flag manifold has positive holomorphic sectional curvature. For the standard homogeneous Kähler-Einstein metric with Einstein constant 2, it is shown that the spectrum of the Calabi operator lies in $(-2,2]$. We determine its sharp upper bound in terms of the anticanonical degrees and characterize the 2-eigenspace. The spectral bounds imply nonnegative cross quadratic bisectional curvature and positive dual cross quadratic bisectional curvature, proving conjectures of Ni and Zheng. We also show that these conditions do not characterize homogeneity: we construct a nonhomogeneous Kähler-Einstein 13-fold whose Calabi spectrum lies strictly inside $(-2,2)$ after the same normalization, as well as nonhomogeneous examples in dimensions 4 and 5. The proofs involve an integral identity for holomorphic sectional curvature, projective embeddings, root-space estimates, and explicit metrics on projective bundles.

comments
57 pages; comments are welcome
fields
Differential Geometry
pdf ↗ page arXiv:2611.07402 ↗

On the projective line, we construct a pair of smooth Kähler potentials whose weak geodesic has isolated quadratic zeros of the fiber form at two fixed points throughout an interior time interval. The geodesic is smooth near these zeros, and its space-time form has constant rank one there. For the same fixed endpoints, the minimum fiber eigenvalue of the nondegenerate Monge-Ampère approximations tends to zero, answering AIMPL Problem 2.5 in the negative. The construction works on every compact connected Riemann surface with a prescribed Kähler form and realizes any positive even order of vanishing. Taking products yields higher-dimensional examples in which the fiber rank drops by one along a smooth hypersurface while the space-time form retains maximal rank nearby. The proof starts from a semipositive model built from a branched cover. We make its endpoints strictly positive by local corrections and use comparison along holomorphic curves to prove that the weak geodesic still agrees with the model on an explicit interior region. There a ramification point of local degree $m$ gives a fiber zero of order exactly $2m-2$.

comments
16 pages
fields
Differential Geometry
pdf ↗ page arXiv:2611.07451 ↗

In this paper, we propose the notion of a \textit{hypermap}, defined as a Riemannian map whose range distribution is tangent to a real hypersurface of a complex space form. This construction leads to a new direction in Riemannian map theory. We establish the fundamental Gauss and Codazzi equations for a hypermap and use them to investigate the extrinsic geometry of its range leaves. In particular, when the ambient complex space form is non-Euclidean, we prove that these leaves cannot be totally umbilical and that their shape operator fails to be parallel. We further incorporate the Hopf condition and define \textit{Hopf hypermap}, thereby obtaining the analog of Hopf hypersurfaces. We also examine the umbilical case and show that an umbilical hypermap cannot exist to a non-Euclidean complex space form. This motivates introducing the \textit{$η$-umbilical hypermap}, which broadens the class of examples for studying the interaction between the range geometry and the ambient complex structure. Finally, we analyze the principal curvatures associated with these hypermaps, including results on their multiplicities and constancy.

fields
Differential Geometry
pdf ↗ page arXiv:2611.07471 ↗

Boundary static potentials arise from the adjoint equations associated with simultaneous variations of scalar curvature and boundary mean curvature. On a connected manifold of dimension $n\ge2$ with nonempty smooth boundary, the space of boundary static potentials has dimension at most $n$. We classify all complete manifolds attaining this bound. For $n\ge3$, every such manifold is isometric to a domain in a simply connected space form and has at most two boundary components. In dimension two, the classification also includes coverings of annuli and exteriors, as well as spherical Möbius bands. Finally, for $n\ge3$, the dimension hypothesis is sharp for constant-curvature rigidity, and for $n\ge2$, it is sharp for the bound on the number of boundary components.

fields
Differential Geometry
pdf ↗ page arXiv:2611.07620 ↗

In this paper, we derive an explicit expression for the Riemann–Christoffel curvature tensor of a trans-Sasakian manifold with pointwise constant $\varphi$-sectional curvature and introduce the notion of trans-Sasakian space forms. As consequences, we obtain explicit formulas for the Ricci tensor and the scalar curvature. We further prove that every $3$-dimensional trans-Sasakian manifold necessarily has pointwise constant $\varphi$-sectional curvature. In addition, we investigate statistical structures on trans-Sasakian manifolds and introduce the notion of trans-Sasakian statistical space forms. We then establish a classification of such statistical structures. Our results show that every trans-Sasakian statistical space form of dimension greater than $3$ is necessarily almost-trivial. In the $3$-dimensional case, we establish a rigidity theorem for trans-Sasakian statistical space forms and provide explicit examples showing the existence of non-trivial $3$-dimensional trans-Sasakian statistical space forms

fields
Differential Geometry
pdf ↗ page arXiv:2611.07680 ↗

For every real Berwald strongly convex complex Finsler manifold $(M,F,J,\hat\nabla)$, if the Cartan nonlinear connection satisfies the parallelism condition $\hat\nabla J=0$, we construct a good complex vertical connection $D$, prove that $D$ is the complex Rund connection, and $F$ is Kähler-Berwald. The inverse direction also can be seen from the construction and proof. As an application, we prove that on an irreducible Hermitian symmetric space $M=G/H$, $G$-invariant complex Finsler metrics are Kähler–Berwald. Besides, we obtain the characterization of strongly pseudoconvex holomorphically invariant complex Finsler metrics on the classical domains of type-IV.

comments
12 pages, no figures. Completed in December 2022
fields
Differential Geometry, Complex Variables
pdf ↗ page arXiv:2611.07752 ↗

In this paper, we study the anisotropic mean curvature flow in the Gaussian space endowed with a Finsler metric obtained from a Minkowski metric by a conformal deformation. We first derive the conformal transformation formulas for the fundamental tensor, the Cartan tensor, and the Chern connection of the conformally related Finsler metrics. Based on these formulas, we express the anisotropic mean curvature in the Gaussian space in terms of the corresponding anisotropic mean curvature with respect to the original Minkowski metric, and derive the associated evolution equation. Finally, we establish the short-time existence and uniqueness of smooth solutions by reformulating the geometric evolution equation as a strictly parabolic quasilinear partial differential equation.

fields
Differential Geometry
pdf ↗ page arXiv:2611.07791 ↗

We give a direct proof of the weak interlacing inequalities \[ω_p\leq\widetildeω_p\leqω_{p+1}\] between the volume and half-volume spectra of a closed connected Riemannian manifold $M^{n+1}$, $n+1\geq3$. We also prove that, for a residual set of smooth metrics, the area of every smooth embedded half-volume CMC boundary lies outside the rational span of the areas of smooth closed embedded minimal hypersurfaces. For dimensions $3\le n+1\le 5$, these results yield strict interlacing for generic metrics and a lower bound of order $A^{n+1}$ for the number of distinct half-volume CMC boundaries with area at most $A$.

fields
Differential Geometry
pdf ↗ page arXiv:2611.07801 ↗

The existence of a transverse metric on a foliated manifold is typically used as a framework for studying transversely elliptic basic differential operators. This article details a method for constructing such a transverse metric directly from the principal symbol of the operator. Consequently, rather than treating the transverse metric as a mere working assumption, we demonstrate that its existence is a necessary condition for the very presence of a transversely elliptic operator.

fields
Differential Geometry
pdf ↗ page arXiv:2611.07903 ↗

We prove that a proper biharmonic surface with flat normal bundle in the four-dimensional unit sphere $\mathbb S^4$ lies minimally in a small hypersphere of radius $1/\sqrt{2}$. Since the surfaces with constant mean curvature and those with parallel normalized mean curvature vector field are already understood, the core of the paper is the non-existence of biharmonic surfaces with flat normal bundle in $\mathbb S^4$ whose mean curvature is not constant and whose normalized mean curvature vector field is not parallel. We first show, adapting an argument of Fu, Hong and Yang, that the gradient of the mean curvature is a principal direction. Then the problem reduces to a first order system of ordinary differential equations with a constraint, and we show that this system has no admissible solutions by an elimination argument.

comments
16 pages
fields
Differential Geometry
pdf ↗ page arXiv:2611.08066 ↗

Let $f:M^n\to\Sn^{n+q}(1)$ be a minimal immersion of a closed connected manifold with $n\ge3$ and $q\ge2$, and let $S$ be the squared length of its second fundamental form. We prove an integral inequality that gives $\max_M S\ge n$ if the immersion is not totally geodesic. If $n\ne4$ or $q=2$, equality $\max_M S=n$ holds if and only if the immersion is a Clifford minimal hypersurface; if $n=4$ and $q\ge3$, equality holds if and only if the immersion is a Clifford minimal hypersurface or the standard minimal embedding of $\mathbb{CP}^2$ in $\Sn^7(1)$. Both classifications are understood up to an ambient isometry, a totally geodesic inclusion, and a finite covering. We construct the integral inequalities using polynomial weights and divergence identities. The required algebraic estimates are proved by sums of squared tensor norms.

comments
34 pages
fields
Differential Geometry
pdf ↗ page arXiv:2611.08126 ↗

This paper provides a classification of analytic actions of the semi-orthogonal groups $\text{SO}^\circ(p,1)$ and $\text{SO}^\circ(p,2)$, for $p \geq 3$, on closed, connected manifolds of dimension $p$ and $p+1$ respectively. It is a sequel to a previous paper, which treats the case $p,q \geq 3$. Adapting Uchida's methods, we explicitly construct analytic actions of $\text{SO}^\circ(p,1)$ on $\text{S}^{p}$ and $\text{S}^{p-1} \times \text{S}^{1}$, and analytic actions of $\text{SO}^\circ(p,2)$ on $\text{S}^{p+1}$, $\text{S}^{p} \times \text{S}^{1}$ and $\text{S}^{p-1} \times \text{S}^{2}$, and we classify them up to analytic isomorphism, using a result of Hitchin and Schneider's classification of analytic $\text{SL}_2(\mathbb{R})$ actions on surfaces. The main result demonstrates that any analytic $\text{SO}^\circ(p,q)$ action, $q=1,2$, on a closed, connected $(p+q-1)$-dimensional manifold is covered by one of the constructed actions or by an action on $\text{SO}^\circ(p,q) \times_P \text{S}^1$, where $P$ is a maximal parabolic subgroup of $\text{SO}^\circ(p,q)$.

fields
Differential Geometry
pdf ↗ page arXiv:2611.08149 ↗

In this paper, we first prove Appleton's conjecuture on the uniqueness of noncollapsed steady solitons constructed in \cite{Appleton}. We also classify non-Ricci-flat $κ$-noncollapsed cohomogeneity one steady gradient Ricci solitons in dimension four and five. After passing to a finite cover, in dimension four, they must be the Bryant soliton or Appleton's noncollapsed 4d steady soliton; in dimension five, they must be the Bryant soliton or solitons constructed by Ivey–Dancer–Wang.

comments
57 pages
fields
Differential Geometry
pdf ↗ page arXiv:2611.08206 ↗

We prove that a connected, properly embedded minimal surface without boundary in $\mathbb{R}^3$ and with density two at infinity is a catenoid or a singly periodic Scherk surface.

comments
89 pages, AI-assisted
fields
Differential Geometry
pdf ↗ page arXiv:2611.08212 ↗

This is the first of a series of papers which will be devoted to a study of the index of mesoscopic Wilson–Dirac operators on closed Riemannian manifolds. The aim of the series is to discretise the index of the Dirac operator. On flat tori, this is achieved by the Wilson–Dirac operator of lattice gauge theory on a regular lattice. On a general closed Riemannian manifold, however, no regular lattice is available. As an intermediate step between the continuum Dirac operator and its discretisations, we introduce mesoscopic Wilson–Dirac operators on any closed Riemannian manifold, defined by integrating covariant differences over distances comparable to a fixed length scale. In this paper, we define the index of the mesoscopic Wilson–Dirac operators via spectral flow and prove that it agrees with that of the continuum Dirac operator at all sufficiently small scales. The proof rests on an elliptic estimate uniform in the scale and a Rellich-type compactness theorem. In the second paper, we shall pass from mesoscopic operators to discrete Wilson–Dirac operators.

fields
Differential Geometry, Mathematical Physics, K-Theory and Homology
pdf ↗ page arXiv:2611.08266 ↗

We prove that every closed weakly stable constant mean curvature surface immersed in the Riemannian Heisenberg space ${\rm Nil}_3$ is an embedded rotational sphere. As a consequence, we solve the isoperimetric problem in ${\rm Nil}_3$: the isoperimetric domains of ${\rm Nil}_3$ are exactly the regions enclosed by rotational constant mean curvature spheres, and for each prescribed volume there is a unique such domain, up to ambient isometry. For small volumes, the isoperimetric result was proved by Tomter in 1993.

comments
12 pages
fields
Differential Geometry, Analysis of PDEs
pdf ↗ page arXiv:2611.08309 ↗

Let $f\colon M^n\to\mathbf{R}^{n+1}$, $n\geq2$, be a smooth immersion of a closed oriented manifold, with Gauss map $ν\colon M\to\mathbf{S}^n$. For each unit vector $u$, the shade $S_u\subset M$ is the set of points where $ν$ makes an acute angle with $u$. We show that $f$ is a convex embedding if and only if every shade is contractible. More generally, if every shade is $k$-connected, then $f$ is a convex embedding provided that $n\leq 2k+1$, or $n=2k+2$ and $H_{k+1}(M;\mathbf{Z})=0$. On the other hand, for every $n\geq 2k+3$ there is a nonconvex embedding $f\colon\mathbf{S}^n\to\mathbf{R}^{n+1}$ all of whose shades are $k$-connected. These results give sharp answers to a question of H. Wente in all dimensions.

comments
7 pages
fields
Differential Geometry, Algebraic Topology
pdf ↗ page arXiv:2611.08415 ↗

We determine the dimension of the space of $\mathfrak{sl}(2)$-invariant (n)-ary differential operators on weighted densities for arbitrary weights, including the singular cases in which the usual elimination procedure breaks down. The proof relies on the strong Lefschetz property established by Stanley \cite{s} and Watanabe \cite{w}. We also investigate the corresponding problem in the super case.

fields
Differential Geometry
pdf ↗ page arXiv:2611.08418 ↗

Discovering continuous symmetries directly from differential-equation data can reveal geometric structure for model reduction and scientific machine learning without symbolic determining equations or a closed-form representation of the governing system. We formulate intrinsic Lie point-symmetry discovery as an operator-theoretic problem on sampled jet-space data. Admissible generators are identified from a sampled determining operator and separated from redundant, solution-dependent, and geometrically trivial directions. The framework combines prolongation, sequential subspace restrictions, and a fixed nonlinear dictionary linear in trainable coefficients. A rank-stabilization preflight test diagnoses systems whose represented intrinsic symmetry algebra does not stabilize under dictionary refinement. On a controlled differential-equation benchmark with an analytically known 11-dimensional represented intrinsic symmetry algebra, the method recovers the complete represented generator set, reproduces its Lie-bracket structure to numerical precision, and approximates a known generator beyond the training domain. Here we show that explicit intrinsic continuous-symmetry generators can be recovered from data obtained from exact or approximate differential systems.

comments
63 pages, 10 figures. Code and data availability statements are included in the manuscript
fields
Differential Geometry, Dynamical Systems, Group Theory, Numerical Analysis
pdf ↗ page arXiv:2611.08421 ↗

The goal of this paper is to investigate the following question: what geometric restrictions does the existence of a transitive family of constant mean curvature hyperspheres impose on a Riemannian metric on $\mathbb S^n$? The notion of a transitive family can be made precise through the theory of Zoll families developed by Ambrozio, Marques, and Neves. Within this framework, we partially extend their local rigidity theorem for even conformal factors on $\mathbb S^3$ to every dimension $n\ge4$, assuming each hypersphere in the minimal Zoll family has the area of a round equator. Moreover, we give a partial answer to the following higher-dimensional version of a question posed by Robert Bryant: must a metric on $\mathbb S^n$, conformal to the round metric and admitting a Zoll family of hyperspheres with a common nonzero constant mean curvature, have constant sectional curvature? Finally, we also investigate inverse problems in the spirit of Hangan, asking which metrics make a fixed family of geodesic spheres have a common prescribed mean curvature.

fields
Differential Geometry
pdf ↗ page arXiv:2611.08423 ↗

In this note, we show that the sum of the Morse index and number of intersections of min-max geodesics achieving the p-widths of a closed surface are bounded above by $p$, improving a result from our previous work. A key idea is that deformations coming from vertex desingularization and eigenfunctions of the Jacobi operator on geodesics are sufficiently independent.

fields
Differential Geometry
pdf ↗ page arXiv:2611.08450 ↗

In this paper, we prove that every complete noncompact two-sided properly embedded constant mean curvature hypersurface of finite Morse index in Euclidean space is minimal. The result holds in every dimension. We assume no bounds on curvature or volume growth. This establishes the properly embedded case of the finite-index version of do Carmo's question. For nonzero mean curvature, we use the radii of inscribed balls to define a positive weight. This weight gives a polynomial bound on weighted volume, while its reciprocal satisfies a differential inequality involving the Jacobi operator. We combine these properties with stability outside a compact set to obtain finite total area. This leads to a contradiction.

comments
13 pages. All comments are welcome!
fields
Differential Geometry, Analysis of PDEs
pdf ↗ page arXiv:2611.08454 ↗

We consider polar foliations of codimension two on compact simply connected irreducible symmetric spaces. In rank at least two, we prove that these foliations are hyperpolar, settling the remaining case of the hyperpolarity problem discussed by Lytchak. We show that, for irreducible spherical Coxeter quotients, the horizontal universal cover fails to be a spherical building precisely on the Cayley plane, where every such foliation is congruent to one of the two known homogeneous models.

comments
19 pages, no figures
fields
Differential Geometry
pdf ↗ page arXiv:2611.08480 ↗

Probability26

The Behrens–Fisher fiducial interval for the difference of two normal means is known from the numerical work of Robinson (1976) to be conservative, but a proof has been lacking. The results here prove that the fiducial interval for any linear combination of $k$ normal means has coverage probability at least its nominal level for all values of the unknown variances. Taraldsen (2026) explains this in more detail using the original data of Behrens (1926). The proof presented in the following rests on a novel plug-in inequality: if a threshold function controls a scaled sum of pivots pointwise in the scale vector, then it also controls the sum with the observed scales plugged in, provided the location errors satisfy a concentration condition formulated through a weighted geometric mean. Independent normal errors satisfy the condition.

fields
Probability, Statistics Theory
pdf ↗ page arXiv:2611.06987 ↗

We give necessary and sufficient coefficient conditions for stochastic exponentials driven by multidimensional diffusions to be true martingales. The classical reduction to non-exit of the diffusion with Girsanov-shifted drift separates the analysis into smooth boundary points, infinity, and nonsmooth boundary pieces. On smooth domains, under regularity and non-explosion hypotheses, martingality is equivalent to nonpositivity of the shifted Fichera drift, provided an explicit local condition holds where that drift vanishes and the stochastic integrand is unbounded. At infinity, Feller's test gives exact criteria when a scalar function of the diffusion reduces to a one-dimensional diffusion up to a time change with rate bounded above and away from zero. General radial and projection tests supply sufficient conditions. Applications give explicit necessary and sufficient conditions for inverse-CIR and $4/2$ models with correlated factors and singular stochastic integrands, including critical equality cases. The coupled non-affine examples go beyond direct scalar Feller tests and affine Riccati criteria, and exhibit boundary exits not decided by tests at infinity. Further exact criteria cover logarithmic boundary singularities, weighted distances, and an absorbed factor. The martingale conditions also yield uniqueness criteria for associated valuation equations.

comments
Keywords: Stochastic exponential; true martingale; strict local martingale; multi-dimensional diffusion; Fichera drift; Feller's test; Khasminskii's criterion; stochastic volatility; affine and Wishart processes; uniqueness for degenerate parabolic equations
fields
Probability
pdf ↗ page arXiv:2611.07083 ↗

We prove a high-probability lower bound for the second eigenvalue in modulus of the transition matrix of the simple random walk on a directed configuration model with minimal out-degree at least two and maximal degree not too large. The lower bound matches the upper bound found by Coste (2021) on the same model. Our proof bypasses the need to establish convergence of the empirical eigenvalue distribution. Instead, it reduces the spectral lower bound to a central limit theorem for the trace of the transition matrix at powers much larger than the logarithmic scale, a range that previous results failed to reach. Both the general strategy and the central limit theorem might be of independent interest.

comments
26 pages, 1 figure
fields
Probability, Combinatorics
pdf ↗ page arXiv:2611.07104 ↗

Many real-world applications, such as numerical weather prediction or wildfire tracking, involve modeling highly nonlinear dynamical systems that are chaotic and influenced by various sources of uncertainty. To improve model prediction accuracy and reduce such uncertainties, data assimilation (DA) combines the dynamical model with partial, noisy observations to estimate the evolution of the system's state. Doing so depends on solving the optimal filtering problem: computing the conditional distribution of a hidden state given a sequence of observations. Classical approaches fail along three distinct axes: Kalman-based methods assume Gaussian posteriors and produce biased estimates under nonlinear observation operators; particle-based methods represent distributions with weighted samples and collapse in high dimensions due to the curse of dimensionality; and kernel-based methods move beyond discrete particle representations but cannot maintain their density approximation affordably as dimension grows. The Ensemble Score Filter (EnSF) avoids all three constraints by encoding the filtering density in a score function and approximating that score with a training-free Monte Carlo estimator, enabling stable filtering in high-dimensional, chaotic systems. This survey traces EnSF's mathematical foundations, establishes it as the state of the art for non-Gaussian, high-dimensional filtering, and examines fourteen extensions that expand its operational reach. Twelve of the fourteen extensions preserve this training-free Monte Carlo estimator at the core of the update, with only two departing from it to achieve planetary-scale deployment.

fields
Probability
pdf ↗ page arXiv:2611.07160 ↗

In the random Lorentz gas, a point tracer moves through spherical obstacles of (reduced) density $Φ$ centered on a Poisson point process in $d$-dimensional space, $\mathbb{R}^d$, and becomes localized once the vacant space stops percolating, at $Φ_{\rm p}(d)$. The dimensional scaling of this void percolation threshold has remained elusive. Prior bounds, conjectural ones included, ranged from $Ω(1)$ to $O(d\ln d)$, leaving even the scaling undetermined. We here prove that $Φ_{\rm p}(d)=Θ(d)$, and in particular that $\tfrac12\,d \le Φ_{\rm p}(d) \le 5\,d$ for all sufficiently large $d$. The upper bound counts hierarchical families of escape waypoints called tree embeddings; the lower bound is via a weighted second-moment estimate on an explicit ensemble of ballistic escape routes. A linear scaling of $Φ_{\rm p}(d)$ excludes the $d\ln d$ growth suggested by analogy with other systems in the random first-order transition universality class and is consistent with the numerical estimate $Φ_{\rm p}/d \to 3.40(5)$. Because the mean-field dynamical transition is likewise linear, $Φ_{\rm d}\simeq 2.40\,d$, our result fixes the activated window $[Φ_{\rm d},Φ_{\rm p}]$ as a bounded—rather than diverging—correction to mean-field dynamical arrest in sufficiently high $d$.

comments
20 pages, 4 figures
fields
Probability
pdf ↗ page arXiv:2611.07196 ↗

We prove weak well-posedness for the scalar stochastic heat equation on the one-dimensional torus with uniformly nondegenerate diffusion that is Holder continuous of any positive order. This resolves the weak uniqueness problem in the full positive Holder range, extending our earlier result for exponents above two thirds (arXiv:2608.01279). The proof couples a weak solution to a smooth companion and uses spatial cancellation to estimate the signed covariance difference in a negative Sobolev norm. A finite Gaussian expansion on each Littlewood-Paley block gives the decay needed to remove the coupling drift by a change of measure. A localized Girsanov argument extends the result to measurable drifts of linear growth.

code
github.com/hanyi162013-Yihan/scalar-she-lean ∎ Lean project (lakefile.toml, lean-toolchain) is too large to build within the check's time limit at 60a6397, checked 2026-11-06 build details
comments
121 pages. Lean formalization: https://github.com/hanyi162013-Yihan/scalar-she-lean
fields
Probability
pdf ↗ page arXiv:2611.07310 ↗

For many probability measures on $\mathbb{R}^n$, and for $k$ sufficiently smaller than $n$, most $k$-dimensional coordinate marginals are close to product measures. We prove this in two regimes, which notably involve no exchangeability assumptions as in prior theorems of de Finetti type. First, for log-concave measures, the quantitative bound depends on the measure only through the operator norm of its correlation matrix, and $k$ may grow polynomially with $n$. A key ingredient of the proof is a new strengthening of the subadditivity of relative entropy for strongly log-concave measures, which as a byproduct yields sharp bounds in the strongly log-concave case. Second, for a broad class of measures satisfying just a variance inequality much weaker than Poincaré's, $k$ must stay below the sharp threshold $2\log n/\log \log n$. Here the basis of $\mathbb{R}^n$ is fixed a priori, which makes this phenomenon quite different from known Gaussian approximations for low-dimensional projections such as the central limit theorem for convex bodies, although there are remarkable parallels. Insofar as Gaussian approximations are driven by the thin shell property, our results are driven by variance inequalities for additive functions which have appeared in the literature on the Gibbs sampler.

fields
Probability
pdf ↗ page arXiv:2611.07460 ↗

We study nonlinear optimal stopping and domineering claims in a Brownian–Poisson filtration, under a nonlinear evaluation induced by a Lipschitz BSDE driver satisfying a comparison condition. A nonlinear domineering claim is a strong $\mathcal{E}^f$-martingale dominating the reward process at every stopping time. We first establish the existence of such claims for every optional reward in $\mathcal{S}^2$. For left-limited, right-upper semicontinuous rewards, we prove a local dual representation at every stopping time: the nonlinear optimal stopping value equals the essential infimum of the values of all domineering claims defined on the remaining stochastic interval. This infimum is attained by a claim constructed from the associated reflected BSDE. We further show that a global domineering claim attains the value at a prescribed stopping time if and only if the reflection processes have accumulated no increase up to that time, with the appropriate endpoint convention. When the reward is right-upper semicontinuous in $\mathcal{E}^f$-expectation and the family of contact times is nonempty, its essential infimum is optimal. Under strict comparison, we also establish a necessary contact condition for optimality. No convexity assumption on the driver is required.

fields
Probability
pdf ↗ page arXiv:2611.07743 ↗

Functional inequalities provide a language for reasoning about the long-time properties of ergodic Markov processes in terms of their macroscopic geometry. This article offers an informal introduction to functional inequalities for readers who are more familiar with probabilistic tools based around couplings, Lyapunov techniques, and similar. The main ideas are developed through the overdamped Langevin diffusion, where various functionals related to this convergence admit particularly transparent expressions, and for which the key associated inequalities -- namely, the Poincaré and logarithmic Sobolev inequalities -- are especially well-studied. We explain the qualitative content of these inequalities, what they imply about convergence to equilibrium, how they transform and compose under natural operations, and how they interface with classical probabilistic approaches. Related inequalities are discussed more briefly, and an annotated guide is provided for readers wishing to enter the wider literature.

comments
21 pages of writing, 4 pages of references, 6 figures
fields
Probability, Functional Analysis
pdf ↗ page arXiv:2611.07785 ↗

We investigate recurrence and uniqueness of continuous-time Markov chains using moments of integral-type functionals. Unlike the classical criteria from Chen (2004) and the Feller-Reuter criterion, we develop general unified conditions for recurrence and uniqueness of Markov processes based on moments of integral-type functionals. The associated criteria follow directly from these conditions whenever the mean functionals can be expressed explicitly. As applications, explicit recurrence and uniqueness criteria are derived for single-birth, single-death, and birth-death processes on trees.

comments
14 pages
fields
Probability
pdf ↗ page arXiv:2611.07821 ↗

We study the asymptotic propagation speed of travelling wave solutions to the stochastic Fisher-KPP equation with multiplicative finite-activity pure jump Lévy noise. The noise is represented by a compensated Poisson random measure with finite intensity, so that the solution evolves deterministically between successive jump times and undergoes multiplicative jumps at the event times. Our main contribution is a stochastic Feynman-Kac representation for the solution, which expresses the random field as the product of the exponential martingale of the jump noise and an expectation over the paths of an independent Brownian motion. This representation is the key probabilistic tool that converts the nonlinear stochastic partial differential equation into an expectation problem, thereby allowing for a direct analysis of the front position. Building on this representation and on the large deviation analysis of the jump noise, we derive explicit upper and lower bounds for the asymptotic wave speed. These bounds are given in terms of the noise intensity $ε$ and the Lévy measure $μ$, providing a quantitative description of how discontinuous stochastic forcing affects the propagation speed in the Fisher-KPP equation. Under the finite-activity assumption, a stopping-time iteration yields pathwise $L^\infty$ bounds and finite-speed propagation of the level sets of the front, which are essential for establishing the existence of a travelling wave profile. The framework extends naturally to multi-species cooperative systems, illustrating its robustness. Numerical simulations are included to demonstrate the informativeness of the theoretical speed bounds.

comments
40 pages, 1 figure
fields
Probability, Analysis of PDEs
pdf ↗ page arXiv:2611.07868 ↗

We prove that, with high probability, the Erdős–Rényi random graph $G(n,p)$ is reconstructible up to isomorphism from the multiset of its rooted induced 2-neighborhoods whenever $p\geq n^{-3/4}(\log n)^3$. Together with the non-reconstructibility result of Johnston, Kronenberg, Roberts and Scott, our result locates the second reconstruction transition at the scale $p=n^{-3/4}$ up to polylogarithmic factors, resolving a question they posed. Our proof relies on a global rigidity argument exploiting 5-cycle constraints.

comments
25 pages
fields
Probability
pdf ↗ page arXiv:2611.07890 ↗

This monograph sets out a constructive theory of stochastic processes in the continuum, in the sense of constructive field theory: the continuum is not postulated but reached as the limit of a sequence of lattice approximations, with the renormalisation the limit requires exhibited rather than assumed, and the limit controlled by estimates that can be checked. The primitive object is not a sigma-algebra but de Finetti's prevision, the price of finitely many gambles at a fixed resolution, which coherence turns into a conditional expectation. The theorem at the centre of the book, called plainly the fundamental theorem, is at once de Finetti's fundamental theorem of probability, the fundamental theorem of Bayesian statistics, and the one-period fundamental theorem of asset pricing; the financial vocabulary is a convenience only. The applications are two, mathematical finance and machine learning, and they are one mathematics read in opposite directions: an assessment is a price in the one and a prediction in the other, and both are computed by one adjoint recursion, the adjoint of a propagation in the first case and backpropagation in the second. Carried through, the two directions give sharp rates of convergence rather than asymptotic orders: for the moments of Ito diffusions, for the value function of Feynman-Kac semigroups, and for the measure changes of the fundamental theorem, the last as a limit theorem for reweighted generators, under a condition keeping the reweighting well behaved. A measure change is thus calculated, not postulated: a lattice approximation with Farkas' lemma computes it, completeness makes it unique, and the passage to the continuum is earned by the regularity of what is computed. Sixty-seven Python scripts regenerate every numerical table, and much of the finite-dimensional core of the argument has been verified in Lean 4.

comments
350 pages, 7 figures. Eleven chapters and a computational appendix reproducing 67 Python scripts, which regenerate every numerical table in the book
fields
Probability
pdf ↗ page arXiv:2611.07944 ↗

In this paper, we construct the complex Ginzburg-Landau dynamic associated with the harmonic oscillator, with a random force which is white in time and covers the full subcritical range in space. We are able to renormalise the non-linearity using the same constant as the one used to renormalise the problem associated to the Laplacian. By an extension of the problem to complex times, we are able to prove that the solution of the complex Ginzburg-Landau is equal in law to the solution of the purely dissipative case, i.e. the associated complex $Φ^4$ model, but along a complex axis of time.

comments
26 pages
fields
Probability, Mathematical Physics, Analysis of PDEs
pdf ↗ page arXiv:2611.08038 ↗

We prove uniqueness in law for three stochastic evolutions on normalised Laguerre-Pólya functions arising as scaling limits of matrix characteristic polynomial dynamics. In particular, this implies that Ahn's universal Brownian Gibbs line ensemble arising from random matrix products has the law of the logarithms of the reciprocal zeros of the solution, unique in law, to a simple stochastic equation on entire functions started from $\mathrm{e}^{-z}$. The method of proof relies on some integrable structure of the problem in the special form of a new duality with finite-dimensional diffusions which is closely related to intertwinings.

comments
23 pages
fields
Probability
pdf ↗ page arXiv:2611.08073 ↗

For a rowwise independent triangular array $\{X_{n,i}:1\le i\le n\}$, put $S_n=\sum_{i=1}^nX_{n,i}$ and $V_n^2=\sum_{i=1}^nX_{n,i}^2$. We prove that $S_n/V_n\xrightarrow{\mathrm d}{\mathrm N}(0,1)$ implies $\max_{i\le n}|X_{n,i}|/V_n\xrightarrow{\mathbb P}0$. The same assumption gives $V_n^2/b_n^2\xrightarrow{\mathbb P}1$ for any choice of medians $b_n^2$ of $V_n^2$. Combined with the results of Shao (2018), this yields necessary and sufficient conditions for the self-normalized central limit theorem and resolves his conjecture for independent sequences. The proof identifies the joint limits of the sum and its sum of squares, then uses the standard normal ratio to remove every non-Gaussian component.

comments
56 pages
fields
Probability
pdf ↗ page arXiv:2611.08137 ↗

The Keller-Segel particle system as defined in \cite{ft} through the Dirichlet form theory (see \cite{f}) consists of $N$ planar Brownian motions interacting through the kernel $x\in \rr^2 \mapsto θK(x) = -θx/\|x\|^2$, where $θ>0$. One particularity of this model is that collisions (i.e. overlapping of particles) occur depending on the value of $θ$, making the well-posedness of the particle system delicate. In this work, we propose a new notion of solution of this SDE without relying on the Dirichlet form theory and prove uniqueness in law of such solutions. The proof relies on an excursion decomposition of the particle system based on the "barycenter-radius-angle" decomposition in \cite{ft}. As additional results, we prove that the solution is strong Markov, has the Feller property and fits in the definition of solutions in the Dirichlet form sense. Finally we prove that the law of the particle system starting from a configuration where all the particles overlap is explicit.

fields
Probability
pdf ↗ page arXiv:2611.08207 ↗

It is well known that for SDEs with non-Lipschitz coefficients, the Euler scheme can fail to converge in an Lp sense. It is also widely observed that convergence fails only due to rare explosive trajectories, that distort the expectation. In this work we demonstrate that almost any method of removing these rare explosive trajectories leads to convergence in a suitable strong and weak sense. We apply this framework to coefficients that grow faster than linearly at infinity and Lennard-Jones type coefficients that explode like a reciprocal power on a compact set.

fields
Probability, Numerical Analysis
pdf ↗ page arXiv:2611.08208 ↗

The Fibonacci spiral is formed by a square-joining rule: begin with a unit square and attach squares whose side lengths follow the Fibonacci sequence; in the standard construction, choices of direction produce a spiral and a tiling of the plane. We study a randomization of this construction, in which the attachment axis alternates deterministically, while fair coin flips determine direction. For a lattice point we ask when the growing rectangle first covers that point and how the expected covering time depends on its location. Tracking the rectangle's boundary reduces this question to random subset sums of even- and odd-indexed Fibonacci numbers. We extend the construction to d dimensions, where coordinate subsequences are super-increasing, allowing us to derive exact covering probabilities and formulas for the expectation and variance. For $d\geq2$, the expected covering time grows logarithmically with distance from the initial block, while the variance remains uniformly bounded. We obtain exponential tail bounds for the delay beyond the earliest feasible covering stage. Finally, we extend the Fibonacci spiral to the randomized tiling through a square-by-square construction using circular arcs and quintic connectors. This construction uses one-step lookahead, admits a continuously differentiable parameterization, and recovers the Fibonacci spiral for the standard tiling.

comments
33 pages, 6 figures
fields
Probability
pdf ↗ page arXiv:2611.08317 ↗

We study Gaussian fluctuations, moderate and large deviations around the high-density hydrodynamic limit of critical and nearly critical space-time Hawkes processes initialized by a prescribed Poisson field of past events, with no subsequent immigration. Under a finite mean reproduction delay and centered displacements with finite spatial covariance, a positive cluster-transform estimate yields a distribution-valued central limit theorem (CLT) and a moderate deviation principle (MDP) throughout the moderate range. For finite nonzero mass of the initial profile, we obtain a full large deviation principle (LDP) for the space-time event measure in the vague topology; its good rate function is a quadratic branching-control energy with a prescribed initial trace and a finite-mass domain. The lower bound uses an explicit approximation by smooth reproduction controls, and the upper bound uses a microscopic exponential martingale and a global generator approximation. We distinguish exact-mean centering from PDE centering and compute deterministic bias corrections under second delay and fourth spatial moments. For a factorized heavy-tail power-law delay, we establish the CLT, MDP, and full measure-valued LDP for the space-time measure, with finite nonzero mass of the initial profile for the LDP. Its fractional control energy retains the source generated by the prescribed past. A source-preserving regularization proves approximation of every finite-action measure. Compact-kernel bounds for initial profiles of infinite mass and the degenerate case of a zero initial profile are stated separately.

comments
72 pages
fields
Probability
pdf ↗ page arXiv:2611.08339 ↗

We study the stochastic Burgers equation on the whole space, driven by space-homogeneous Gaussian noises that are white in time and colored in space, for both positive and zero viscosity. We show that in dimension one, for any fixed average velocity, there is a unique ergodic global solution to the Burgers equation for all nonnegative viscosities, and that the One Force—One Solution principle holds. We also describe the basin of attraction and show that these global solutions are one-point pullback attractors. Moreover, the viscous global solution converges to the inviscid global solution as the viscosity tends to zero. Our technique relies on the study of polymer measures; accordingly, we establish the existence and uniqueness of infinite-volume polymer measures in a given asymptotic direction, as well as their convergence to semi-infinite minimizers in the zero-temperature limit. We also establish the existence part of all these results in arbitrary dimension larger than one.

fields
Probability
pdf ↗ page arXiv:2611.08355 ↗

We prove a logarithmic central limit theorem for a tagged particle in infinite-dimensional Dyson Brownian motion at inverse temperature $β=2$, starting from the sine$_2$ point process. The fluctuation is of order $\sqrt{\log t}$, which is subdiffusive and substantially slower than the polynomial fluctuations of classical short-range interacting particle systems, reflecting the rigidity induced by long-range logarithmic repulsion. We further identify the scaling limit as the singular Gaussian field corresponding to the zero-Hurst parameter, which provides a realisation of the singular Gaussian field as a scaling limit of a tagged particle raised by Dürr, Goldstein and Lebowitz (Comm. Pure Appl. Math. 38 (1985)).

comments
Comments welcome!
fields
Probability, Mathematical Physics
pdf ↗ page arXiv:2611.08365 ↗

The geometry of spherical spin glasses is relatively well understood, thanks in large part to the Kac-Rice formula. However, this formula does not directly apply to discrete spin systems, and much less is known mathematically about the geometry of Ising spin glasses. In this paper, we present a simple method for counting local optima and gapped states in a broad class of discrete spin glasses. To illustrate its scope, we compute the annealed complexities ("configurational entropies," in the physics language) of mixed Ising spin glasses with potentials, such as external fields, and of the Ghatak-Sherrington model. These computations recover several existing results in the mathematics and physics literature as special cases. We can also compute energy-constrained complexity, finding a critical energy threshold analogous to the spherical case, but for gapped states of pure Ising spin glasses. Finally, we use our method to establish a form of disorder persistence for gapped states, identifying a transition between $3$-spin and $4$-spin glasses.

comments
57 pages, zero figures
fields
Probability, Mathematical Physics
pdf ↗ page arXiv:2611.08374 ↗

We develop a new framework for sampling at low temperatures using Glauber dynamics. In this regime, the coexistence of competing modes (commonly referred to as phases) creates bottlenecks that can make worst-case mixing exponentially slow. The main new ingredient in our framework is to relate convergence within a phase to the geometry of disagreements between two independent configurations drawn from that phase. As applications of our approach, we consider three canonical models on random regular graphs that have been central to low-temperature sampling, and establish fast convergence from natural initialisations of Glauber dynamics for every fixed degree. We show: --- For the hard-core model on random bipartite $d$-regular graphs, Glauber dynamics initialised from either one-sided independent set mixes within the corresponding phase in $n^{3/2+o(1)}$ steps throughout the non-uniqueness regime $λ>λ_c(d)$. --- For the random-cluster model on random $d$-regular graphs, for every real $q>2$, Glauber dynamics initialised from the all-in edge set mixes within the ordered phase in $n^{3/2+o(1)}$ steps for all $β\geq β_c$. Similarly, Glauber dynamics initialised from the all-out edge set mixes within the disordered phase in $n^{3/2+o(1)}$ steps for $β\leq β_c$. --- For the ferromagnetic Ising model on random $d$-regular graphs, Glauber dynamics initialised from all-plus or all-minus mixes within the corresponding phase in $n^{3/2+o(1)}$ steps throughout the low-temperature regime $β>β_c(d)$. We also give an $n^{2+o(1)}$-time algorithm that combines phase samples into samples from the full distribution, avoiding in particular time-costly approximations of the corresponding partition functions to the target sampling accuracy.

comments
Abstract shortened to fit arXiv requirements
fields
Probability
pdf ↗ page arXiv:2611.08416 ↗

We consider vectors chosen randomly from high-dimensional hypersurfaces of the form $\{x\in\mathbb{R}^n: \sum_{j=1}^n V_j(x_j)=an\}$. These in general asymmetric surfaces generalize classical spheres and include unbounded and disconnected surfaces. We show that under mild assumptions as $n\to\infty$, any weighted sum of the random coordinates is asymptotically Gaussian, provided that no weight dominates. An explicit reduction in the variance compared to the unconstrained case is observed. We also study a model with penalization of deviations from the constraint, interpolating between the unconstrained and the constrained case. Our approach involves an explicit representation of the characteristic function of the constrained random variables in terms of independent random variables.

fields
Probability
pdf ↗ page arXiv:2611.08446 ↗

Gu and Zhang asked for almost sure upper and lower bounds on the Taylor coefficients $a_n$ of $\exp(\sum_{k\ge1}X_kz^k/\sqrt{k})$ in the unit disc, and for the dependence of these bounds on the input tail. For independent, identically distributed, rotationally invariant complex inputs with $\mathbb E|X_1|^2=1$, we prove the universal almost sure bound $|a_n|=O((\log n)^{1/4})$. Almost surely, $\limsup_{n\to\infty}|a_n|/(\log n)^{1/4}$ is positive and finite exactly when $\mathbb E[|X_1|^2\log^+|X_1|^2]<\infty$, and is zero otherwise. For each input law, we define a deterministic scale such that, almost surely, the coefficients are bounded exactly when the scale is bounded, and tend to zero exactly when it vanishes. We also determine the almost sure logarithmic growth exponent as the maximum of contributions from collective fluctuations and individual large inputs. Explicit tails exhibit distinct thresholds for boundedness and for upper bounds at the scale in the phase criteria. Examples show that neither the leading tail nor the set of finite logarithmic moments determines the almost sure phase.

comments
65 pages
fields
Probability
pdf ↗ page arXiv:2611.08467 ↗

Number Theory18

In this paper, we provide an upper bound on the number of a special system of Roth-type Diophantine inequalities over proper adelic curves. By refining the technique from previous work, we successfully eliminate the large-height restriction on the solutions. Furthermore, the constraints on the parameters within the inequality are further relaxed. We also give a constructive quantitative generalization of the Vojta-Dolce-Zucconi system, replacing compactness-based extraction by a discrete pigeonhole argument and a finite grid partition, so that the resulting bound has explicit dependence on the relevant parameters.

comments
22 pages, comments are welcome!
fields
Number Theory
pdf ↗ page arXiv:2611.06973 ↗

Extending a result of Kocik, we show that the tangency points of the Ford circles built over quotients of $k$-Fibonacci numbers are concyclic: they lie on a circle centred on the horizontal axis, of radius $\sqrt{k^2+4}/2$, meeting it at the reciprocal of the metallic ratio $σ_k$ and at $-σ_k$. Concyclicity turns out to be a quadratic identity of the sequence read in coordinates. In the Poincaré half-plane, this circle is a geodesic along which the tangency points advance with constant step $2\lnσ_k$.

comments
10 pages, 5 figures
fields
Number Theory
pdf ↗ page arXiv:2611.07141 ↗

In 1984, Mahler asked which analytic functions map the set $\mathscr L$ of Liouville numbers into itself. In the first paper of this series, we proved that every real-analytic solution on an open interval is the restriction of a real rational function. We now identify a canonical subfield $\mathbb L$ of $\mathbb R$ such that, for every nonconstant $R\in\mathbb R(X)$ and every nonempty open interval $I$ in its domain, $R(I\cap\mathscr L)\subseteq\mathscr L$ if and only if $R$ is defined over $\mathbb L$ and $R^{-1}(\mathbb Q)\cap I\cap\mathscr L=\varnothing$. The weaker inclusion $R(I\cap\mathscr L)\subseteq\mathbb Q\cup\mathscr L$ holds if and only if $R$ is defined over $\mathbb L$. If $R$ is not defined over $\mathbb L$, every such interval contains a Cantor set of Liouville numbers whose images under $R$ are irrational and have irrationality exponents bounded by a constant depending only on $R$. We also construct continuum many positive strong Liouville numbers $σ$ for which $σX^m$ preserves $\mathscr L$ for every $m\geq2$, with all nonzero rational fibers consisting of $U_m$-numbers.

comments
31 pages, Parte II of the complete solution to Mahler 1984 problem on Liouville numbers
fields
Number Theory
pdf ↗ page arXiv:2611.07177 ↗

We develop a geometric rigidity theory for Mahler measure based on the global structure of Ronkin functions and amoebas. Our starting point is the theorem of Ronkin describing the affine linearity of the Ronkin function on the complement of the amoeba and the identification of its gradients with lattice points of the Newton polytope. Combining this deep analytic input with convex duality and toric geometry, we show that the vanishing of the Mahler measure is an extremal phenomenon: it occurs if and only if the Ronkin function collapses to the piecewise linear support function of the Newton polytope. This collapse forces a degeneration of the amoeba onto the $(n\!-\!1)$-skeleton of a toric fan and implies that the zero locus of the polynomial is contained in a translate of a codimension-one subtorus. As a first application of this rigidity principle, we recover the generalized Kronecker theorem characterizing Laurent polynomials of vanishing Mahler measure. Our approach bypasses the classical Boyd–Lawton limit formulas and replaces analytic projection arguments by a global convex-geometric and potential-theoretic mechanism rooted in the geometry of amoebas.

fields
Number Theory, Complex Variables
pdf ↗ page arXiv:2611.07181 ↗

We study the finite extensions of a number field generated by a single point on an elliptic curve. We construct geometrically non-isomorphic elliptic curves with complex multiplication that generate the same collection of extensions, giving a negative answer to a question of Mazur and Rubin. For torsion-generated fields, we prove that inclusion of the collections up to finitely many exceptions implies isogeny over the ground field, and in the non-CM case even isomorphism. When all geometric endomorphisms are defined over the ground field, we characterize equality of the torsion-generated collections by the existence, for every prime $p$, of an isogeny over the ground field of degree prime to $p$. This condition also implies equality of the full collections of fields generated by a single point.

comments
Comments welcome!
fields
Number Theory
pdf ↗ page arXiv:2611.07259 ↗

For an integer $b\neq0,1$ let $K_b$ be the cubic field generated by the unique real root of $x^3-3bx-b^3$, let $η_0$ be its fundamental unit, and let $η_b$ be the explicit unit studied in the companion paper (arXiv:2609.37389). Write $η_b=η_0^{n(b)}$. Deciding whether $n(b)$ is divisible by $q$ amounts to finding the integral points of a plane curve $C_q$ of degree $2q$. We prove that for every integer $q\ge1$ the curve $C_q$ is irreducible with $g(C_q)=\binom{q-1}{2}$, the genus of a smooth plane curve of degree $q$ rather than of degree $2q$. The proof is not a genus computation but a structural one: $C_q$ is the inverse image, under the $q$-power map of a two-dimensional torus, of the curve traced by the minimal polynomial of $η_b$ as $b$ varies, and the formula falls out of two applications of Riemann-Hurwitz in the resulting tower. As a consequence the problem splits into exactly three regimes: $q=2$ rational, $q=3$ elliptic, and $q\ge4$ of general type. In particular only finitely many $b$ have $q\mid n(b)$ for each $q\ge4$. A lower bound for linear forms in two logarithms gives $n(b)<6082138$ for every $b$, so $n(b)\in\{1,2\}$ for all but finitely many $b$, though not effectively. We add two computations: $n(b)$ is determined outright for $|b|\le10^{30}$, and the rational points of $C_4$ are determined by quadratic Chabauty, giving $4\mid n(b)\iff b\in\{-75,-3,5,96\}$. We state precisely which parts of the latter rest on certified arithmetic and which do not. Finally, we show that $K_b\cong K_{b'}$ with $b\neq b'$ holds for $\{b,b'\}=\{-1,5\}$, and otherwise only if one of the two parameters, $c$ say, satisfies $|c|>10^{30}$ and $n(c)$ is divisible by a prime $q\ge5$. There are only finitely many such pairs.

comments
30 pages. PARI/GP and SageMath scripts: https://github.com/iwaokimura/cubic-fields-unit-powers (v1.2)
fields
Number Theory, Algebraic Geometry
pdf ↗ page arXiv:2611.07333 ↗

Let $b\ge2$ be an integer and let $D\subset\mathbb{N}_0$ be a finite digit set containing zero, with maximum residue multiplicity $μ=\max_{0\le r<b}\#\{d\in D:d\equiv r\pmod b\}$ and cardinality $q=\#D\ge2$. Under the condition $μ<q$, we prove that the summatory representation function satisfies an asymptotic formula with a strictly positive, Hölder continuous, $1$-periodic fluctuating profile and an error of $O_D(x^{\log_bμ})$. An exact periodic factorization of the Laplace generating function furnishes an explicit, computable Fourier expansion for the periodic profile via Gamma factors, and yields the meromorphic continuation of the associated Dirichlet series to the entire complex plane. When $D$ is a complete residue system modulo $b$, we establish an arithmetic synchronization criterion showing that the associated finite carry system admits a zero-synchronizing word if and only if $\gcd(D)=1$. This synchronization yields an exponentially accurate interval approximation and a general transfer principle for arithmetic weights. Consequently, we derive a prime number theorem for representable primes governed by the same periodic profile, alongside an exponentially convergent measurable scaling limit of the representable sets.

comments
20 pages. This paper extends the analytic results of arXiv:2305.00792 by establishing a synchronization criterion for the finite carry system and deriving a prime number theorem for representable numbers
fields
Number Theory, Dynamical Systems
pdf ↗ page arXiv:2611.07446 ↗

Adopting notation from Erdős [Congress. Numer., 1980], we define $w_1(n) = w(n)$ for positive integers $n$ so that $w(1) = 0$ and so that: If $n > 1$ and $n = \prod p^{α_p}$, then $w(n) = \sum_{p^{α_p} \parallel n} \frac{n}{p^{α_p}}$. By then defining $V(x)=\bigl|\{w(n):1\le n\le x\}\bigr|$ for $x \geq 1$, Erdős considered whether or not $V(x) = o(x)$ and left this as an open problem. This same problem was also given as an open problem in Guy's famous textbook on unsolved problems in number theory and seems to have remained open. We solve this problem by proving that $V(x) \neq o(x)$, and, moreover, by proving the stronger result such that $\liminf_{x \to \infty} \frac{V(x)}{x} > 0$.

comments
Submitted for publication
fields
Number Theory
pdf ↗ page arXiv:2611.07617 ↗

Congruence jumping replaces coordinates by integer quotients and updates polynomial divisibility relations. We extend this procedure to several variables and retain integer coefficients by recording their common factor separately. For three-variable systems of divisibilities with signed binomial numerators, we classify all positive integer exponent tuples of period at most two under three successive coordinate jumps: seven fixed tuples and twenty-two cycles of exact period two. No odd period greater than one occurs. At the chosen representatives of the twenty-two exponent two-cycles, there are eighty-four returning sign assignments. All have realizations with nonzero pairwise-coprime integer coordinates, and exactly seventy-four have such realizations with every coordinate positive throughout the six jumps. Explicit constructions and congruence arguments establish this classification. We also study jumps that decrease a specified height, distinguishing points that admit no immediate decrease from minima of entire connected components of the jump graph. For the four-variable family $Q_λ(x,y,z,w)=kxyzw$, where $Q_λ$ is the sum of the four squares plus $λ$ times the six pairwise products, finite bounds give complete classifications for $λ=1,2,3$. For $λ=1$, positive integer solutions exist exactly for $k=2,8,10$. These results distinguish returns of the polynomial data from integrality and positivity of the coordinate sequences.

fields
Number Theory, Dynamical Systems
pdf ↗ page arXiv:2611.07625 ↗

We study a genus-two hypergeometric family whose Jacobians have real multiplication by $\Q(\sqrt5)$. We identify the relevant primitive factor and integral rank-two monodromy; at one real embedding the projective image is $Δ(2,5,\infty)$, while the conjugate image is nondiscrete. At the primes above $11$, the degree-$12$ Borel covers are quadratic pullbacks of genus-one triangular modular curves. The same rank-two systems have finite-field hypergeometric Frobenius traces, yielding the local Euler factors and a quadratic transformation twisted by the character of $2$.

comments
22 pages
fields
Number Theory, Algebraic Geometry
pdf ↗ page arXiv:2611.07849 ↗

The minimal excludant (mex) of an integer partition $λ$ is the smallest positive integer that is not a part of $λ$. In this paper we study an analogue of this statistic for $\ell$-regular partitions with $\ell\geq2$, that is, partitions in which no part is divisible by $\ell$. Recently, Meher \cite{Meher} defined the minimal excludant for $\ell$-regular partitions. Motivated by Meher \cite{Meher}, for an $\ell$-regular partition we define its least $r$-gap as the smallest positive integer not divisible by $\ell$ that appears fewer than $r$ times as a part, and we obtain the two-variable generating function for the number of $\ell$-regular partitions of $n$ with a given least $r$-gap. Let $σ^{\mex}_{\ell}(n)$, and $σ^{\mex}_{\ell,r}(n)$ be the sum of the of the mex values over all $\ell$-regular partitions of $n$ and least $r$-gaps over all $\ell$-regular partitions of $n$ respectively. Using Jacobi's triple product identity, we derive explicit expressions for the generating function of $σ^{\mex}_{\ell}(n)$, and $σ^{\mex}_{\ell,r}(n)$, with separate formulas for even and odd $\ell$. We also show that the sequence $\{σ^{\mex}_{\ell,r}(n)\}_{n\ge0}$ is weakly increasing for every $\ell\ge2$ and $r\ge1$. Also, using Ingham's Tauberian theorem, we establish Hardy–Ramanujan type asymptotic formulas for $σ^{\mex}_{\ell,r}(n)$ when $\ell=4,5,7$ and $r\ge1$ is arbitrary. Lastly, we prove $σ^{\mex}_{\ell,r}$ is odd infinitely often, and we establish a quantitative lower bound, showing that the number of such n up to X grows at least like log log X.

comments
22 pages, first draft of the paper
fields
Number Theory
pdf ↗ page arXiv:2611.07871 ↗

We prove a finiteness theorem for Dirichlet series satisfying the Euler product, analytic continuation, and functional equation expected of a holomorphic newform. For any fixed level $N$, even weight $k$, and exponent $σ$, only finitely many such series have Dirichlet coefficients $\ll n^σ$. No hypotheses on twists by Dirichlet characters are assumed. The proof combines a combinatorial spanning theorem for Ramanujan sums with an operator-theoretic argument that uses two distinct primes to establish discreteness of the possible Hecke eigenvalues.

comments
32 pages
fields
Number Theory
pdf ↗ page arXiv:2611.07921 ↗

In this article, we study the Mordell Weil ranks of elliptic curves in the family $E_D:y^2=x^3+Dx$, where $D\in\mathbb{Z}\setminus\{0\}$. Under suitable congruence conditions on $D$ and its prime divisors, we use $2$-descent via a rational $2$-isogeny to study the rank of $E_D$. Under our congruence hypotheses, the ranks of these matrices give sufficient conditions for $E_D$ to have rank zero. We give explicit sufficient conditions for the curves $E_D$ to have Mordell Weil rank zero over $\mathbb{Q}$ in the following cases: (i) $D$ is an odd negative integer, (ii) $D$ is an odd positive integer, and (iii) $D$ is a positive even integer. Also we give a quantitative result describing how many square free parameters satisfy the required matrix-rank condition. In particular, our results show that these rank zero criteria hold for a significant proportion of the admissible parameters.

comments
15 pages. Comments, suggestions, and feedback are most welcome
fields
Number Theory
pdf ↗ page arXiv:2611.08029 ↗

We prove that $K$-valued overconvergent $F$-isocrystals over smooth $k((t))$-varieties are overholonomic and construct the triangulated category of overholonomic complexes of Frobenius type which is stable under six operations and contains overconvergent $F$-isocrsytals.

fields
Number Theory
pdf ↗ page arXiv:2611.08193 ↗

We prove an arithmetic doubling identity for the standard arithmetic theta series on $X_0(N)$ at arbitrary positive level, with constant $ψ(N)/24$ for the fixed genus-two Eisenstein series. The product comparison covers all pairs of integral indices. At rank zero, an explicit Epstein zeta correction agrees with the middle Bruhat orbit of the Eisenstein derivative. The common-component calculation combines integral CM adjunction, the cyclic difference divisor, and compact Green-current identities. It retains the different, conductor, internal contacts, and proportionality contacts at every prime, including dyadic and supersingular places. Finite intersections, curvature, and discriminant heights identify the nonpositive coefficients with an explicit modular completion. Uniform estimates for parameter derivatives give strong Fourier convergence. Green-energy estimates justify the double Petersson pairing. The full cusp lift exists as a Bochner integral and equals its Mordell–Weil projection. For $\mathbb Z/(2N)\mathbb Z$, explicit prime-power recursions determine the complete zero-component trace spectrum, including the dyadic factor. They give the scalar-to-vector Petersson norm correction and an isometric normalization on the active subspace. The Bruinier–Yang theorem then yields the scalar height formula on the specified Jacobi newform line. We also prove a double Rankin–Selberg criterion based on positive diagonal coefficients and record the paired analytic factorization with its ramified local integrals retained explicitly.

comments
151 pages
fields
Number Theory
pdf ↗ page arXiv:2611.08243 ↗

We study the set of visible points from the origin in certain cut-and-project sets, specifically the density of the set. We generalize to all real quadratic number fields the formulas for the density found by G. Hammarhjelm, and the corresponding error bounds proved by I. Gringlaz, R. Kumar, B. Weiss. We also find the possible densities of the set of visible points in the cut-and-project set, with a constant generic lattice, and a changing window with constant volume. We also study bounds for the distance of large holes in the set of visible points of certain cut-and-project sets, generalizing theorems originally proved by P. Erdős for $\mathbb{Z}^2$.

fields
Number Theory, Dynamical Systems
pdf ↗ page arXiv:2611.08257 ↗

Let $p$ be an odd prime, and let $V_1$ and $V_2$ be symplectic self-dual $p$-adic de Rham representations of $G_{\mathbb{Q}}$ that are pure (of weight $-1$) and satisfy the Panchishkin condition at $p$. Suppose that $V_1$ and $V_2$ are residually isomorphic, compatibly with their Panchishkin filtrations. We prove the relative $p$-parity conjecture relating the ratio of the global root numbers of $V_1$ and $V_2$ to the difference of their Bloch-Kato Selmer dimensions. We have no restriction on the dimension of the Galois representation, the Hodge-Tate weights or the monodromy operator at $p$, and the residual isomorphism need not be symplectic. The proof expresses ratios of local root numbers in terms of local invariants over the cyclotomic $\mathbb{Z}_p$-extension: a new invariant introduced at $p$, and local Iwasawa $λ$-invariants at $\ell\neq p$. Both are compatible with the arithmetic local constants of Mazur-Rubin and Nekovář. A key input, proved using Sen theory, is that the Bloch-Kato and strict local conditions at $p$ coincide over the cyclotomic $\mathbb{Z}_p$-extension, generalising results of Coates-Greenberg and Nekovář. As applications, we obtain new cases of the $p$-parity conjecture, including cases of rank greater than two: Artin twists of CM Hida families, orthogonal Artin twists of ordinary modular forms, symmetric cubes along Hida families, and families of abelian surfaces, together with unconditional numerical examples.

comments
37 pages
fields
Number Theory
pdf ↗ page arXiv:2611.08316 ↗

We prove that K3 surfaces with no rational points over the finite field with $q$ elements exist if and only if $q \leq 11$. Our approach applies linear programming techniques to improve on generic sphere packing bounds for the discriminants of even rootless definite lattices. These bounds are applied to constrain the possible zeta functions of K3 surfaces with no rational points via the Artin–Tate formula. The remaining cases are ruled out by direct analysis of primitive embeddings of particular lattices into Niemeier lattices.

comments
20 pages
fields
Number Theory, Algebraic Geometry
pdf ↗ page arXiv:2611.08356 ↗

Numerical Analysis18

Efficient radiation transport on modern high-performance computers requires algorithms that expose spatial parallelism while maintaining robust convergence. One-cell inversion (OCI) replaces the spatially sequential transport sweep of source iteration with a cell-local solve, making discrete ordinates transport space-parallel, but its spectral radius approaches unity in diffusive problems, and existing accelerators such as transport synthetic acceleration reintroduce sweeps. Here we show that projecting the OCI residual onto a low-order angular space on cell boundaries accelerates convergence in scattering-dominated problems without a transport sweep. We implement the projection in a two-dimensional simple-corner-balance discretization with two angular bases: a linear ($P_1$) basis and one augmented with a quadrant moment ($P_1Q$). Fourier analysis of an infinite homogeneous medium shows that both corrections reduce the spectral radius in optically thick, highly scattering regimes, with $P_{1}Q$ being most effective. Across several test problems, including a modified version of the void-duct streaming benchmark, the method accelerates diffusive and scattering cases but does not uniformly improve convergence in streaming dominated problems. We also find that $P_{1}Q$ is a more robust accelerator than $P_1$ in situations where the flux has strong angular variation. Boundary projection substantially improves OCI convergence in diffusive regimes, and these results delimit the regimes that require additional stabilization or richer projection spaces.

comments
28 pages, 17 figures, 2 appendices
fields
Numerical Analysis
pdf ↗ page arXiv:2611.06951 ↗

The numerical simulation of kinetic plasma models in six-dimensional phase space poses significant computational and memory challenges, particularly when considering modern high-performance computing architectures. In this work, we propose an adaptive dynamical low-rank approach for the efficient solution of the 6D Vlasov–Poisson system. Our strategy is an extension of the projector-splitting dynamical low-rank integrator. The algorithm changes the rank based on physical observables (e.g., the electric field and energy) as opposed to controlling the error in the distribution function. Our approach is memory efficient in the sense that only rank $r$ objects are required to implement the scheme. This makes the algorithm suitable for performing high-dimensional computations on graphic processing units. Comparisons with rank adaptive Basis-Update & Galerkin integrators exhibit both significant memory savings and reduction in run time. We show, by presenting a variety of 6D benchmark experiments implemented in the C++/CUDA low-rank framework Ensign, that our technique allows us to run simulations with a significantly smaller rank, while still preserving the physical fidelity. Thus, the approach substantially reduces memory footprint and computational cost and enables simulations that are prohibitively expensive for full-rank or less memory-efficient low-rank schemes.

fields
Numerical Analysis
pdf ↗ page arXiv:2611.07107 ↗

In the lattice Boltzmann method (LBM), the momentum-exchange method (MEM) is widely used to evaluate hydrodynamic forces at fluid-solid interfaces due to its computational efficiency and geometric flexibility. However, when mapping complex geometries onto structured Cartesian grids, the physical boundary rarely intersects the lattice links exactly at their midpoints. This discrepancy violates an implicit assumption inherent to the conventional momentum-exchange (CME) formulation. In this study, we systematically analyze this geometric deficiency and demonstrate that it induces significant force inaccuracies, even in elementary flows where the solid boundary deviates from the midway position. To resolve this, we propose a modified momentum-exchange (MME) method that explicitly accounts for the exact boundary location along the lattice links, demonstrating enhanced accuracy for both flat and curved geometries. Furthermore, we address another critical issue with the interpolation-supplemented bounce-back (ISBB) schemes, and that is their inherent violation of local mass conservation. This mass leakage induces unphysical pressure buildup in the system, which compromises the accuracy and stability of the LBM. We propose a mass-conservative remedy for ISBB schemes that eliminates this leakage. Through a series of benchmark simulations of increasing complexity, we demonstrate that combining the proposed MME with the mass-corrected ISBB scheme drastically improves the fidelity of hydrodynamic force evaluations, establishing a robust framework essential for moving boundary problems and fluid-structure interactions.

comments
lattice Boltzmann methods
fields
Numerical Analysis
pdf ↗ page arXiv:2611.07137 ↗

Rational solutions of partial differential equations (PDEs) are notoriously difficult to approximate via spectral Fourier methods due to their algebraically slow decay rate. In this work we discuss approximating rational PDE solutions in a basis of orthogonal functions known as the Christov functions, allowing for the computation of its spectrum via the fast Fourier transform. An explicit fourth-order integrating factor method is derived. The effectiveness of the method is tested on several nonlinear PDEs. Relative to a straightforward pseudo-spectral approximation, the integrating factor method supports larger regions of stability resulting in a more efficient scheme.

fields
Numerical Analysis
pdf ↗ page arXiv:2611.07154 ↗

We consider numerical approximations of shallow-water and thin-film flow models with capillary effects. These models lead to third-order differential equations, and we work with an augmented model with second-order derivatives. For a wide class of such models, which includes the classical shallow-water model and thin-film flow models, we develop a fully discrete energy-conserving scheme for both the quadratic and fully nonlinear capillary models. Numerical results show that the proposed scheme is stable and conserves total energy in the absence of dissipative effects.

fields
Numerical Analysis
pdf ↗ page arXiv:2611.07360 ↗

Kinetic equations in the high-field regime describe particle transport dominated by electric-field acceleration and collisional relaxation. These equations are widely used in simulations of plasmas and semiconductor devices. Typical models include the Vlasov–Poisson–Fokker–Planck (VPFP) and Boltzmann–Poisson systems. Their high dimensionality and multiscale nature pose two of the major challenges to accurate and efficient numerical solution. To address these challenges, we propose an asymptotic-preserving tensor neural network (APTNN) method. Specifically, to reduce high-dimensional computational costs and memory requirements, we adopt the canonical polyadic (CP) decomposition structure of TNNs to approximate their solutions. We show that the equation residuals retain this decomposition form, allowing high-dimensional loss integrals to be evaluated through combinations of one-dimensional integrals. The resulting computational and storage costs are reduced from exponential to polynomial growth with dimension. To improve the prediction accuracy in the high-field limit, we enhance the baseline AP loss by incorporating the first moment equation. The added loss term directly constrains the limiting flux closure to reduce its approximation error in the high-field regime. The enhanced AP loss can still be evaluated dimension-wise. Numerical results show that APTNN improves prediction accuracy in the high-field regime with significantly reduced computational costs and memory usage. For a 3D3V VPFP problem with multiple random inputs, training takes less than 30 minutes on a single GPU.

fields
Numerical Analysis
pdf ↗ page arXiv:2611.07364 ↗

We study randomly shifted rank-1 lattice rules for computing $\mathbb{E}[g(\mathbf{X}_T)]$, where $\mathbf{X}_T$ is the terminal state of a stochastic differential equation (SDE) with constant diffusion matrix. Discretizing the SDE by the Euler–Maruyama scheme with $m$ uniform time steps turns the expectation into an integral whose dimension grows with $m$ and whose integrand has a nested composition structure; both complicate the lattice rule error analysis. We generate the Brownian path by the Brownian bridge construction and handle the nested structure using the multivariate Faà di Bruno formula and a rooted-tree expansion. Under suitable regularity and growth conditions, we prove that an $N$-point randomly shifted lattice rule with a uniformly random generating vector achieves a mean absolute sampling error of $O(N^{-1+\varepsilon})$ for every $\varepsilon\in(0,1/2)$ and every prime $N$, with the implied constant independent of $m$. To improve stability and control higher error moments, the median lattice estimator is used. The estimator takes the median of $R=2k-1$ independent lattice estimates and achieves an $L^q$ error of $O(N^{-1+\varepsilon})$ for every $1\le q<k/(1-\varepsilon)$. We further establish error bounds independent of both the state dimension and the Brownian path dimension under an additional summability condition on the columns of the diffusion matrix. Numerical experiments show that median lattice estimators converge faster than Monte Carlo estimators and remain stable as the number of time steps $m$ grows. For models satisfying the additional summability condition, stability is also observed as the state and Brownian path dimensions increase.

fields
Numerical Analysis
pdf ↗ page arXiv:2611.07424 ↗

We propose a hybrid Newton-BFGS high-index saddle dynamics (HiSD) method for accelerating saddle-point searches. The method constructs a Newton correction on the unstable subspace from the available negative eigenpairs, while BFGS updates based on projected secant information approximate the inverse Hessian on the stable subspace. By combining Newton and quasi-Newton corrections, the method incorporates curvature information in both subspaces while preserving the ascent-descent structure of HiSD. Accounting for the evolving stable subspaces, we establish local $Q$-superlinear convergence of the full-memory version near nondegenerate saddle points under suitable assumptions. We further develop a limited-memory version for large-scale problems and provide fixed-step and line search strategies. Numerical experiments show local convergence behavior consistent with the theory and demonstrate acceleration on ill-conditioned problems and degenerate saddle manifolds in neural-network loss landscapes. The method also supports efficient upward and downward saddle searches for a liquid crystal model, further illustrating its broad applicability to saddle-point computation.

fields
Numerical Analysis
pdf ↗ page arXiv:2611.07692 ↗

We present an axisymmetric continuous simplex space-time finite element method for incompressible fluid flow. Compared to conventional time-stepping methods, the space-time approach has several advantages. First, the motion of the domain is incorporated into the space-time geometry, allowing for large boundary motion without remeshing or projection of the solution. Second, the method allows for time-periodic solutions to be computed directly, without the need to simulate multiple cycles. Third, the space-time mesh can be partitioned in time as well as in space, which enables parallelism in time. The method is demonstrated on an intra-aortic balloon pump (IABP) in a rigid tube. The balloon inflation is modelled based on measured data, and includes a travelling wave along the balloon. The method is able to resolve complex flow features such as jets and recirculation zones. A refinement study shows convergence under temporal, spatial and combined space-time refinement. In weak scaling tests up to 960 ranks, the cost per solver iteration remains approximately constant, and partitioning the mesh into time slabs improves the convergence of the linear solver compared to partitions that cut through both space and time.

comments
20 pages
fields
Numerical Analysis
pdf ↗ page arXiv:2611.07822 ↗

We prove three results about Runge–Kutta methods with non-negative coefficients. First, every $s$-stage implicit Runge–Kutta method of order at least two has SSP coefficient at most $2s$; the bound is sharp and is attained by a known diagonally implicit method, so full implicitness gives no advantage over DIRK at second order. This extends the second-order DIRK result of Kocsis \& Németh to full coefficient matrices and settles a conjecture of Ketcheson, Macdonald & Gottlieb. These proofs were found with the assistance of AI tools, which uncovered an interesting relationship between the Butcher matrices for SSP methods and phase-type distributions, and led to the proof of the second result: a bound on the SSP coefficient of $s$-stage implicit Runge–Kutta method of order at least three. The third result in this work is a proof that a third-order Runge–Kutta method with entrywise non-negative $A$ and $b$ cannot be algebraically stable. This algebraically intensive proof was also AI-assisted. This paper describes the iterative process of AI-assisted results, and presents an approach that centers human understanding and creativity in the AI-assisted mathematical endeavor.

fields
Numerical Analysis
pdf ↗ page arXiv:2611.07981 ↗

In this paper, we propose a boundary-preserving randomized neural network method for Poisson equations with Dirichlet boundary conditions in two and three dimensions. A linear lifting imposes the boundary conditions exactly, without balancing penalty weights between interior and boundary residuals. We construct the liftings by generalized barycentric coordinates and transfinite interpolation on convex and nonconvex polygons and polyhedra, and by radial parameterization on star-shaped curved domains. Numerical experiments on different domains show that the proposed method substantially reduces errors compared with the standard randomized neural network method.

fields
Numerical Analysis
pdf ↗ page arXiv:2611.07985 ↗

Operator learning offers an efficient surrogate for lithography simulation, replacing costly numerical solutions with direct model inference. However, existing learning architectures are not designed to simultaneously provide geometric equivariance and adaptability to layout windows with different physical extents, and thus their generalization capability is limited. To address this problem, we propose LithoSAGE, a spatially adaptive and geometrically equivariant neural operator method trained on lithography data. Within its basic Fourier neural operator architecture, spectral weights are expressed as continuous functions of normalized radial frequency through a learnable sum of Gaussians. This representation preserves the geometric properties of the lithography system and incorporates rotation and translation equivariance, spatial interaction decay, and direct evaluation on different layout windows. Experiments predict aerial images from mask layouts and resist images from aerial images, demonstrating the geometric equivariance and spatial adaptivity of LithoSAGE. Comparisons with other architectures also assess prediction errors and parameter efficiency.

fields
Numerical Analysis
pdf ↗ page arXiv:2611.08001 ↗

Nonlinear low-rank decompositions are receiving growing attention as a way to represent structured data with fewer parameters than conventional linear methods. Existing nonlinear approaches have focused mainly on matrices, which limits their ability to capture multidimensional structure. We propose a nonlinear tensor decomposition based on sigmoid nonlinearization in the Tucker and tensor-train formats. A binary tensor is represented by a low-rank real-valued score tensor, whose parameters are estimated by minimizing the binary cross-entropy loss. The resulting optimization problem is jointly nonconvex, but becomes convex with respect to one factor or score block when the remaining factors are fixed, as in the corresponding matrix case. We evaluate the method on several structured binary tensors and compare it with conventional Tucker and tensor-train decompositions across a range of common ranks. These results indicate that sigmoid nonlinearization is particularly effective in the low-rank regime, where it can provide more accurate binary reconstructions than conventional linear tensor decompositions at comparable model sizes.

fields
Numerical Analysis
pdf ↗ page arXiv:2611.08020 ↗

We propose a second-order, semi-implicit time-stepping scheme for the shallow-water equations on the sphere discretised with compatible finite elements. The scheme splits the advection operator from the wave operator, applying a DG-upwind SSPRK3 method to the advection part and an $H(\text{div})$-conforming implicit discretisation to the wave part, solved efficiently via hybridisation. The scheme follows a three-step symmetric splitting: an explicit half-step, a full advection step, and an implicit half-step that simultaneously projects the velocity back into the $H(\text{div})$ space. This symmetry guarantees global second-order accuracy and confines the stability restriction to the CFL condition of the advection step alone. The time-step can therefore be extended beyond this limit by subcycling the advection step without additional implicit solves. The method is implemented in Firedrake and validated against a standard suite of spherical shallow-water benchmarks, demonstrating stable results for Courant numbers up to 1 (with two advection subcycles) with second-order finite elements, second-order convergence, and good parallel scaling.

fields
Numerical Analysis
pdf ↗ page arXiv:2611.08169 ↗

We prove that Schumaker's lower bound gives the dimension of $S_{3r+1}^r(Δ)$ for every nondegenerate triangulation $Δ$ of a polygonal disk in $\mathbb{R}^2$ and every integer $r\ge3$. After accounting for the local vertex contributions, we reduce the remaining dimension defect to compatible vertex functionals. Apolarity associates a unique polynomial with each face, and compatibility forces differences of adjacent face polynomials to contain powers of both endpoint forms. We eliminate the leading residual coefficient using Hermite relations when $r$ is odd and an additional coefficient comparison when $r$ is even. A uniform induction then raises the endpoint powers until all face differences vanish; the zero boundary face potentials imply that every vertex functional vanishes and the global defect is zero. Together with the known cases $r=1,2$ and Hong's higher-degree theorem, this yields the dimension formula for all $r\ge1$ and $d\ge3r+1$.

fields
Numerical Analysis
pdf ↗ page arXiv:2611.08178 ↗

The Characteristic Mapping (CM) method is a transport scheme which discretizes the Lagrangian flow map as a composition of spatial interpolants. This adaptive approximation technique has demonstrated novel resolution properties and the ability to efficiently simulate the fine scales of inviscid turbulent flows. In this work, we construct an extension of this approach to the solution of dissipative problems, including the Navier-Stokes equations. The extension lifts a general scalar evolution equation to a path in a group of diffeomorphisms, where the semigroup approach of the CM method can be exploited. We give an interpretation of this lift through the geometric framework of optimal transport. We then introduce a semidirect product evolution technique to perform the time stepping of the map components. The map is evolved based on the semi-Lagrangian CM approach, and its log-Jacobian is evolved explicitly by an exponential integrator. The numerical scheme is demonstrated to be high-order accurate and is tested on a range of canonical nonlinear dissipative equations. By retaining the resolution properties of the CM method, the approach shows a promising ability to simulate turbulent fluid flows across all Reynolds numbers.

fields
Numerical Analysis
pdf ↗ page arXiv:2611.08261 ↗

A hybrid virtual element formulation (HVEM) is presented for transient flow in variably saturated soils governed by Richards' equation. The nodal heads are the only global degrees of freedom, and a Darcy flux is assumed in each polygonal element and condensed. The flux combines a divergence-free part with a particular field of linear divergence that absorbs storage and sources, so that the water balance of every element holds identically, at every iterate and for any time step, and that of the whole domain at convergence. To guarantee rank-sufficiency with no stabilisation parameter, the divergence-free fields are harmonic polynomials ordered by their behaviour under rotation, giving a closed-form rule for the number of flux parameters, derived on regular polygons and verified numerically on distorted ones and on elements with hanging nodes. Darcy's law is written in each element for the Kirchhoff potential, which removes the conductivity from the diffusion while the head stays continuous across material interfaces. A Kirchhoff mean of the conductivity makes hydrostatic equilibrium and gravity drainage exact on any mesh. The storage is lumped on the nodes with weights supplied by the assumed flux. All the element operators reduce to boundary integrals, so that the nonlinear iterations need only nodal values. The formulation is assessed on solutions in closed form, on published benchmarks and on rainfall over homogeneous and layered slopes, where independent solutions provide a further check. It converges at second order in the head on polygonal and locally refined meshes, also where gravity dominates, and closes the water budget in every test. On a layered slope, cut grids refined in a band along the surface reach the accuracy of a uniform grid twice as fine with about 40% of the nodes. Its structure, shared with the HVEM for elastoplasticity, suits coupled analyses of slopes.

fields
Numerical Analysis
pdf ↗ page arXiv:2611.08276 ↗

We determine field-of-values equivalent preconditioners for the time-dependent Stokes equations discretized in time by stage-weighted Radau IIA Runge–Kutta time-stepping and discontinuous Galerkin time-stepping methods. By weighing each stage of the Radau IIA Runge–Kutta method differently we obtain preconditioners that are field-of-values equivalent, with constants independent of viscosity, timestep, and mesh size, to the system matrix. Similarly, by using weighted inner products in time to define the discontinuous Galerkin time-stepping method, we again obtain field-of-values equivalent preconditioners for the resulting system matrix. Numerical experiments are performed in two and three-dimensions which demonstrate robustness of the proposed preconditioners.

fields
Numerical Analysis
pdf ↗ page arXiv:2611.08429 ↗

Algebraic Geometry12

We introduce the $(m,p,q)$-Lefschetz property for $(1,1)$-type endomorphisms of Hodge–Lefschetz modules, providing a framework for studying nef $(1,1)$-classes and characteristic classes of nef vector bundles on compact Kähler manifolds. Under a Hall–Rado type condition on nef twisted vector bundles, we prove Hodge–Riemann relations for products of top Chern classes and, in suitable bidegrees, for products of derived Schur classes. For products of nef $(1,1)$-classes, we show furthermore that the Hall–Rado type condition is equivalent to simultaneous Hodge–Riemann relations. Our proofs build on the purity theorem and descent lemma of Cattani, Kaplan, and Schmid. We extend the descent lemma to Schur classes of nef vector bundles. We also establish formal analogs for projective bundle modules over Hodge–Lefschetz modules, with applications to tautological classes and descent on matroid Chow rings.

comments
52 pages
fields
Algebraic Geometry, Combinatorics, Differential Geometry
pdf ↗ page arXiv:2610.37650 ↗

We develop effective calculation tools for constructing categorified invariants of tangles and knots, following the work of Cautis and Kamnitzer. These invariants, originally constructed by means of Fourier-Mukai transform functors between derived categories of coherent sheaves on certain iterated flag varieties, can be given an alternative algebraic description by using tilting bundles. We use this method and provide several examples of its application, in the hope of showing its utility to explicit pen-and-paper computations. This tilting-bundle method also allows us to lift the construction from derived to homotopy categories of complexes of modules, reducing the calculation of tangle invariants to a type of categorified matrix algebra, where the matrices in question have maps between their elements and calculations get turbocharged with chain homotopy.

comments
53 pages, 2 figures. Comments are welcome
fields
Algebraic Geometry, Quantum Algebra
pdf ↗ page arXiv:2611.07046 ↗

We introduce a class of interesting algebraic varieties --- odd biisotropic symplectic Grassmannians $\operatorname{S_2Gr}(k,2n+1)$ and study them with an emphasis on the case of the submaximal odd biisotropic symplectic Grassmannians $\operatorname{S_2Gr}(n,2n+1)$, which we call catalecticant varieties. The main result of the paper is a construction of a full exceptional collection in the derived category of the catalecticant sixfold $\operatorname{S_2Gr}(3,7)$.

comments
33 pages
fields
Algebraic Geometry
pdf ↗ page arXiv:2611.07079 ↗

For a diagonal matrix $D$ with distinct diagonal entries in an algebraically closed field $K$ we study the symmetric matrices $Q$ with entries in $K$ such that the affine line $D+tQ$ has the same eigenvalues in $K$ as $D$. We study the equality of characteristic polynomials, $χ_{D+tQ}(x)=χ_D(x) \text{ for all }\: t\in K$ and examine the resulting system of equations. For generic diagonal matrices of size at most $4$ we show that only the trivial symmetric matrix $Q=0$ satisfies this isospectral condition. We use this result to compute the degree of a field extension that appears in the invariant theory of quadratic forms under congruence action.

fields
Algebraic Geometry
pdf ↗ page arXiv:2611.07150 ↗

We study the passage of special gaps from a Zariski semiroot to a plane branch. A decomposition into rectangular blocks allows us to compute the maximal valuations of the coefficients of differentials and describe the image of the transition map. We treat vertical and horizontal decompositions, as well as decompositions with alternations, obtaining criteria for the preservation of special gaps. For towers whose ramification indices after the first are all equal to two, the value semigroup of the branch and the special gaps of the first semiroot determine the special gaps of the branch.

fields
Algebraic Geometry
pdf ↗ page arXiv:2611.07255 ↗

Let $k$ be a field of characteristic zero with $a,b,c\in k^\times$ and let $m\in k\setminus\{0,4abc\}$. We compute the Brauer group of the generalised Markoff surface $ax^2+by^2+cz^2-xyz=m$. We also construct several counterexamples to the integral Hasse principle, which are explained by the Brauer–Manin obstruction.

comments
32 pages
fields
Algebraic Geometry, Number Theory
pdf ↗ page arXiv:2611.07454 ↗

The theory of wonderful compactifications provides a common framework encompassing the constructions of important varieties such as the Fulton-MacPherson configuration spaces, the moduli space of $n$-pointed stable rational curves, and the De Concini-Procesi models of subspace arrangements, and, more generally, of smooth compactifications in which the subvarieties in a building set are replaced by a simple normal crossings boundary divisor. We give a presentation of the Chow ring of a wonderful compactification, assuming only that the Chow ring of the ambient variety surjects onto that of each subvariety in the induced arrangement. Our results recover and extend many previously known presentations and apply to a broad range of new examples.

comments
49 pages
fields
Algebraic Geometry
pdf ↗ page arXiv:2611.07507 ↗

We determine all roots and their multiplicities in the Bernstein–Sato polynomial of every generic Pfaffian ideal. In the nonmaximal case, the roots lie in explicit rational windows indexed by the rank strata. Integral roots retain their label multiplicities, whereas every nonintegral root is simple, including values with several labels. The proof combines graph-module relations, skew Capelli operators, central characters, and a weighted Mellin integral. We deduce the strong motivic monodromy conjecture for Pfaffian ideals, with pole orders bounded by Bernstein–Sato root multiplicities. For Pfaffians of size $2r$, the largest possible root multiplicity as the matrix size varies is $\lfloor(\sqrt{8r+1}-1)/2\rfloor$. In the nonmaximal case we also prove that the nonunipotent part of the Verdier specialization has zero nilpotent monodromy. Keywords. Bernstein-Sato polynomial, Pfaffian ideal, root multiplicities, strong monodromy conjecture.

comments
34 pages
fields
Algebraic Geometry
pdf ↗ page arXiv:2611.07571 ↗

Let $F_n$ be the Fermat curve of degree $n$ over $\mathbb Q$, and let $D_n=F_n/S_3$ be its quotient by coordinate permutations. We prove that $D_n$ has trivial geometric automorphism group for every $n\ge 11$ with $\gcd(n,6)=1$. Its Jacobian has complex multiplication. For $n=11$, this Jacobian is absolutely simple of dimension five. We construct an isogeny over $\mathbb Q$ of degree $11^5$ from $\operatorname{Jac}(D_{11})$ to the Jacobian of $y^{11}=x(1-x)^2$. For primes $p\equiv 1\pmod 3$, we also consider the metacyclic quotients $C_p$ of $F_p$ over $\mathbb Q$. By a theorem of Carocca, Lange and Rodríguez, their Jacobians are absolutely simple and have complex multiplication. We prove that $C_p$ has trivial geometric automorphism group if and only if $p\ge 19$. Write $p=r^2+rs+s^2$ with $r,s$ positive integers. The natural degree-three cover $Y_p\to C_p$ satisfies $\operatorname{gon}(Y_p)=r+s$, so $\operatorname{gon}(C_p)\ge \lceil(r+s)/3\rceil$. Since $g(C_p)=(p-1)/6$, both the genus and the gonality of $C_p$ tend to infinity with $p$.

fields
Algebraic Geometry, Number Theory
pdf ↗ page arXiv:2611.08176 ↗

We study the specialization of torsion points in one-parameter degenerations of elliptic curves with singular fibers of Kodaira type. We describe the resulting torsion curves, their ramifications, intersection multiplicities, and irreducible and connected components.

comments
35 pages, 0 figures. Comments welcome!
fields
Algebraic Geometry
pdf ↗ page arXiv:2611.08190 ↗

Butler's conjectures predict that, on a general smooth curve $C$ of genus $g\ge3$, a general coherent system $[\mathcal{O}_C\otimes V\xrightarrow{\varphi}E]$ with semistable underlying bundle $E$ and $\dim V>\operatorname{rk}(E)$ has surjective evaluation map $\varphi$ and stable kernel. In this paper, we prove Butler's conjectures for $g\ge8$ in the two cases in which the Brill-Noether geometry of a general curve is known, namely rank one and the range $μ(E)>2g-2$. By deformation and openness, it suffices to establish the result on a single Brill-Noether-Petri general curve. We do so for curves on K3 surfaces of Picard rank one, using the wall-crossing theory of coherent systems developed by the first two authors.

comments
33 pages. Comments welcome!
fields
Algebraic Geometry
pdf ↗ page arXiv:2611.08333 ↗

We study kappa classes on the normalization of the main component of the Kollár–Shepherd-Barron–Alexeev(KSBA) moduli space of stable toric pairs. Using a toric model obtained by a lattice modification, we compute all kappa classes on the moduli space by giving a combinatorial formula for their intersection numbers with all torus-invariant cycles. In particular, we identify the first normalized kappa class with the GKZ polarization of the toric variety corresponding to the secondary polytope. For interval configurations, we also obtain a closed formula for all top intersection products of normalized kappa classes.

fields
Algebraic Geometry
pdf ↗ page arXiv:2611.08337 ↗

Dynamical Systems12

We prove that a transitive monotone factor of a smooth surface diffeomorphism has a unique measure of maximal entropy whenever its topological entropy is positive. Its entropy is the entropy of a single homoclinic class of the diffeomorphism, and the factor map is a measure-theoretic isomorphism on every ergodic measure with positive entropy of that class. The other homoclinic classes are mapped to periodic orbits. We give an application to the essential factor of a diffeomorphism isotopic to a pseudo-Anosov map ---possibly relative to a finite invariant set--- constructed by García-Sassi \& Tal: the essential factor has a unique measure of maximal entropy, which is Bernoulli and has full support.

comments
9 pages
fields
Dynamical Systems, Geometric Topology
pdf ↗ page arXiv:2611.07236 ↗

A prescribed finite-dimensional analysis of an evolving discrete history need not admit an exact recurrence driven only by its current state and the entering input. We study exact realizations of causal linear discrete history operators (DHOs). Operator factorization and kernel inclusion characterize exact closure, while the cross-projection defect characterizes closure failure for orthonormal analyses. For finite-dimensional history spaces over fixed finite horizons, we distinguish two minimality problems. Reachable task quotients determine the minimum state dimension for preserving prescribed future tasks on reachable histories. Backward invariant envelopes characterize the minimum dimension of exact linear realizations recovering prescribed analyses on full history spaces within the admissible class. The minimum number of additional independent state directions is given by the rank of the propagated closure defect operator restricted to the kernel of the current analysis. For geometrically weighted countable histories, matrix mode transport yields structured exact recurrences and a uniform exponential decay estimate for state transitions. Additional conditions provide uniform Gram bounds and refined decay estimates. A stationary Meixner specialization yields explicit recurrence matrices, an O(N) arithmetic state update, and minimum augmentation dimensions over finite horizons for selected coefficients. Numerical experiments examine recurrence accuracy, state transition bounds, Gram conditions, and minimum realization dimensions.

fields
Dynamical Systems
pdf ↗ page arXiv:2611.07396 ↗

How often does an element of the geometric iterated Galois group of a Lattès map fix a vertex of the tree? Let A/K be an abelian variety over a separably closed field of characteristic 0, G <= Aut(A,0) finite, and alpha an isogeny of degree at least 2 normalising G. The Lattès map on X=A/G has a preimage tree governed by alpha-power torsion; we determine the fixed-point proportion in every dimension. For zeta in G the weight is the reciprocal of the number of fixed points of zeta in alpha-power torsion (zero if infinite); equivalently, it is a product of p-adic absolute values of determinants on the alpha-active subspaces; for scalar alpha=m the proportion is (1/|G|) times the sum over 1 != zeta in G of the product over p dividing m of the p-adic absolute value of det(1 - zeta | H_1(A,Z)). We deduce positivity and vanishing criteria, prove the groups G_infinity are martingale groups, determine the limit law, and reduce groups with translations to the fixed-origin case. In dimension one we determine ten values for scalar multipliers and nontrivial G <= Aut(E,0), and for the eight quadratic Lattès classes the proportions are 1/8, 1/4 and 5/16. Bridy-Jones-Kelsey-Lodge exclude the Lattès case from their main theorem; their Theorem 2.5 periodic-point bound is attained, via reductions modulo split primes p = 1 (mod 4), for the map (x^2+a)/(x^2-a), a^2 = -1, and we settle the exceptional case of the Fariña-Asategui-Radi conjecture.

fields
Dynamical Systems
pdf ↗ page arXiv:2611.07816 ↗

This paper deals with the real Thurston pullback algorithm for entire maps with finitely many finite singular values, all real. Given a finite forward-invariant marking containing the singular values and satisfying an interval no-Levy-cycle condition, the pullback is shown to have a unique fixed point and it is shown that every orbit converges to it. The resulting entire map uniquely realises the marked dynamics up to positive affine conjugacy. Thus this paper established Thurston realisation and uniqueness in the setting of real entire maps with a finite postsingular set. As usual, the realisation problem reduces to the existence of a fixed point of the Thurston map $T$ which in the real setting acts on a simplex $W$ of dimension $s-2$ if the required postsingular set consists of $s$ points. The key estimate concerns a finite real preimage interval: when its critical multiplicity is bounded and its endpoint values stay separated from the singular values unless they equal to one, its hyperbolic diameter grows at most double-logarithmically as the separation tends to zero. An expansive marking (absence of Levy cycles) then rules out configurations near the boundary of the simplex whose displacement under the Thurston map is at most a fixed multiple of their distance to the boundary. Using a version of the Brouwer fixed point theorem, this immediately implies that the Thurston map $T$ has a fixed point. The argument permits infinitely many real and non-real critical points and different behaviours as the real variable tends to $+\infty$ and $-\infty$. Additional estimates for fixed labelled inverse branches, which are not strictly necessary for the proof of the Main Theorem, are given in the appendices.

fields
Dynamical Systems
pdf ↗ page arXiv:2611.07818 ↗

For each fixed $g\ge3$, we prove that non-uniquely ergodic directions have Hausdorff dimension $1/2$ on almost every area-one surface in $\mathcal H^{\mathrm{odd}}(g-1,g-1)$ when $g$ is odd, and in $\mathcal H^{\mathrm{nonhyp}}(g-1,g-1)$ when $g$ is even. We construct these directions through cyclic splittings into $g$ slit tori. Effective equidistribution of Lindenstrauss–Mohammadi–Yang supplies sufficiently many children at each normal input. We estimate cusp and periodic-orbit failures with the original branching weights and prove that their total mass loss is summable. The surviving path measure yields the dimension lower bound, also for directions nonergodic for flat area.

comments
25 pages
fields
Dynamical Systems
pdf ↗ page arXiv:2611.07833 ↗

A generalized Hopf bifurcation takes place when varying a parameter of a (semi)dynamical system causes an attractor $K$ to become a repeller. In doing so, it expels a (parameter-dependent) family of attractors. In this paper we concentrate on two questions: (1) what do the expelled attractors look like?, and (2) can any attractor $K$ undergo a generalized Hopf bifurcation? Regarding (1), we compute the cohomology of the expelled attractors in terms of the cohomology of $K$; in the case of continuous-time dynamics this can be improved to a computation of their Borsuk shape. As for (2), we show that certain compacta cannot undergo generalized Hopf bifurcations and, when the ambient space has dimension $ \leq 3$, give a complete topological characterization of which can.

comments
30 pages, 7 figures
fields
Dynamical Systems, Geometric Topology
pdf ↗ page arXiv:2611.08002 ↗

Let $X$ be a separable locally compact metric space, let $E\neq\{0\}$ be a normed space, and let $G$ be a semigroup of continuous self-maps of $X$. Every $\varphi\in G$ induces the composition operator $C_\varphi f=f\circ\varphi$ on the space $C(X,E)$ of continuous functions endowed with the compact-open topology. We characterize the topological transitivity of the semigroup $\widehat{G}=\{C_\varphi:\varphi\in G\}$ by a condition that combines the run-away property of $G$ with an approximate form of injectivity, and we show that, when $E$ is separable, $\widehat{G}$ is topologically transitive if and only if it is hypercyclic. In particular, if every element of $G$ is one-to-one and $G$ is run-away, then $\widehat{G}$ is topologically transitive, and hypercyclic when $E$ is separable. For a single map $\varphi$ we obtain: $C_\varphi$ is hypercyclic $\iff$ $C_\varphi$ is topologically transitive $\iff$ $\varphi$ is one-to-one and run-away, and $C_\varphi$ is mixing $\iff$ $\varphi$ is one-to-one and strongly run-away. If, moreover, $\varphi$ is an open map (which is automatic on manifolds by invariance of domain), then $C_\varphi$ is Devaney chaotic. On an open interval all of these properties are equivalent to $\varphi$ being one-to-one without fixed points.

comments
This is a revised and extended version of a published article
journal
M. Javaheri, Hypercyclic and topologically transitive semigroups of composition operators, Topology Appl. 164 (2014), 105--112
fields
Dynamical Systems
pdf ↗ page arXiv:2611.08320 ↗

Two-stage contagion, in which adoption requires two successive contacts with adopters, combined with population turnover produces bistability and relaxation oscillations in Katriel's differential-equation model, but no chaos. We derive from the elementary reactions of the process, without reference to the differential equations, a discrete-time model in which the population is updated in rounds of length $τ$, as in elections, meetings or turns of a game: each individual makes at most one transition per round, events compete as exponential risks, and adoptions take effect from the next round. The model is a Markov chain whose large-population limit is a map $F_τ$; Katriel's equations are recovered as $τ\to0$. Taking the probability $p$ of adoption after a single contact as bifurcation parameter, with rates and round length fixed, $F_τ$ goes from extinction through a window of chaos to a stable equilibrium; for $β/γ=46$, $μ/γ=0.002$, $γτ=5.5$ the window is $0.023\le p\le0.056$, bounded by the invasion threshold and by a boundary crisis beyond a subcritical Neimark–Sacker point. One-stage contagion ($p=1$), $p=0$ and permanent adoption showed no chaos. The chaotic attractor is essentially unique, with one positive Lyapunov exponent and an outbreak–recharge geometry. In the same window Katriel's equations have a periodic relaxation cycle with troughs of $10^{-44}$--$10^{-164}$: the round structure does not create the oscillation but turns it into chaos and shrinks the oscillatory range in $p$ from $(0.022,0.41)$ to $(0.022,0.056)$. The chaos is therefore not an artefact of discretising a stable state, but a property of two-stage contagion with synchronous updating.

comments
24 pages, 11 figures, 7 tables; includes Supplementary Material (6 pages, 6 figures, 3 tables)
fields
Dynamical Systems
pdf ↗ page arXiv:2611.08353 ↗

We analyze a delay differential model for phage therapy in which a cocktail of $n$ different types of bacteriophages, each with its own latency period, is administered and coinfection is allowed. We prove the global existence, uniqueness and nonnegativity of solutions, and the existence of an invariant region, under suitable conditions, in which the bacteria are eradicated exponentially fast. We determine the stability of the bacteria-free equilibrium in terms of the administered doses and give sufficient conditions for the nonexistence of coexistence equilibria, showing that coinfection may allow bacteria and phages to coexist even when the doses are large. For a cocktail of two phage types with the same latency period, we obtain a complete description of the biologically feasible coexistence equilibria, which may be zero, one or two. Numerical simulations illustrate the theoretical results.

fields
Dynamical Systems
pdf ↗ page arXiv:2611.08370 ↗

Building on our previous work, in which geometric Gibbs measures were introduced to complete the space of classical Gibbs measures under the Teichmüller metric, we establish the uniqueness of geometric Gibbs measures in this paper. This result completes the construction of a Teichmüller complex manifold structure on the space of all geometric Gibbs measures. Using the welding technique, we uncover a connection between geometric Gibbs measures and harmonic measures on dynamical quasicircles. This connection allows us to establish asymptotic conformal rigidity for these dynamical quasicircles. Finally, we establish a foundation for the systematic study of the ergodic properties of geometric Gibbs measures.

comments
47
fields
Dynamical Systems, Mathematical Physics, Complex Variables
pdf ↗ page arXiv:2611.08384 ↗

We study the quantity $\sup_{φ\inΦ}\limsup_{n\to\infty}φ(F_n)/|F_n|$, where $\mathcal F= (F_n)_{n\in\mathbb{N}}$ is a sequence of finite subsets of a countable amenable group, and $Φ$ is a collection of subadditive functions. Under mild conditions on $Φ$ we prove that the aforementioned quantity equals the supremum over $φ\inΦ$ of the limit of $φ(F)/|F|$ as $F$ becomes more and more invariant, multiplied by a ``Følnerness'' constant $C(\mathcal F)\in[1,\infty]$. We have $C(\mathcal F)=1$ if and only if $\mathcal F$ is Følner. A corollary of this result is that every choice of a non-Følner sequence overestimates topological and measure-theoretic entropy by the multiplicative factor $C(\mathcal F)$ (in the positive entropy case). We obtain a generalization of classical results of Goodman, Krug, and Newton, on the so-called proportionality principle for sequence entropy, and their extension by Gao, Huang, and Zhu to countable amenable groups.

comments
17 pages
fields
Dynamical Systems, Group Theory
pdf ↗ page arXiv:2611.08408 ↗

We establish characterizations of h-dichotomies for noninvertible evolution families acting on Banach spaces, in terms of appropriate notions of expansiveness and uniform noncriticality. Our results extend the corresponding characterizations previously obtained for invertible evolution families. To the best of our knowledge, no such characterizations have previously been established for general noninvertible evolution families. A key difference is that noninvertibility makes the structure of the stable spaces more delicate: in our positive-half-line result, finite codimension must be imposed at every time, rather than only at the initial time. We show by an example that the latter, weaker condition is insufficient. We also study evolution families on the negative half-line, using duality to obtain corresponding characterizations in terms of dual expansiveness and dual uniform noncriticality. General growth rates are treated by a time-rescaling argument.

fields
Dynamical Systems
pdf ↗ page arXiv:2611.08451 ↗

Functional Analysis10

Given an orthogonal decomposition \(\widetilde{\mathfrak{H}} = \mathfrak{H}[\dotplus]\widehat{\mathfrak{H}}\) of a Krein space, we study compressions \({\mathrm{C}_{\mathfrak H}}(R)\) of linear relations \(R\) in \(\widetilde{\mathfrak{H}}\) to \(\mathfrak{H}\). We formulate a sufficient condition that ensures that compression commutes with taking adjoints, that is, the adjoint identity \({\mathrm{C}_{\mathfrak H}}(R^{\widetilde{*}}) = {\mathrm{C}_{\mathfrak H}}(R)^{*}\) holds. We derive preservation and reflection results for the compression of maximal dissipative, maximal symmetric, and self-adjoint linear relations. For nonnegative linear relations, the adjoint identity also implies that compression commutes with taking the Friedrichs extension. For \(\mathfrak{H}\ne\{0\}\), we characterize finite dimensionality of the exit space \(\widehat{\mathfrak{H}}\) by universal compression properties. Finally, we give a criterion for the adjoint identity to hold in terms of bounded linear functionals on \(R\).

comments
25 pages
fields
Functional Analysis
pdf ↗ page arXiv:2611.07299 ↗

Let $L$ generate a symmetric contraction semigroup. We prove a holomorphic Hörmander calculus on $L^p(X;\ell^q)$ at the optimal angle $\max\{\arcsin|1-2/p|,\arcsin|1-2/q|\}$, with boundary Sobolev order $s>3$.

comments
29pages
fields
Functional Analysis
pdf ↗ page arXiv:2611.07430 ↗

Yamazaki proved that the norms of the iterated Aluthge transforms of an operator converge to its spectral radius, and Ando showed that for matrices the singular values of the iterates converge to the moduli of the eigenvalues. We prove that the latter holds for every compact operator on a Hilbert space and every generalized Aluthge transform $Δ_t$ with $0<t<1$: for every $j$, the $j$-th singular value of the $n$-th iterate converges to $|λ_j(T)|$, and the products of the $k$ largest singular values, as well as the Ky Fan norms, decrease monotonically to their limits. This is the Aluthge counterpart of the Yamamoto–Davis theorem. We also determine the limit in the Schatten $p$-norm whenever $T$ belongs to a strictly smaller Schatten class.

fields
Functional Analysis
pdf ↗ page arXiv:2611.07728 ↗

We investigate whether a dual Banach space with the approximation property (AP) must necessarily have the bounded approximation property (BAP). Using a quantitative renorming theorem for the space of absolutely summable sequences, we establish a direct-sum construction yielding a negative answer. More precisely, we construct a separable Banach space (X) such that both (X) and its dual (X^*) have the approximation property, while neither space has the bounded approximation property. The construction uses a sequence of equivalent renormings of (\ell_1) with increasingly restrictive finite-dimensional approximation bounds. By assembling these spaces through a (c_0)-direct sum, we obtain a dual space whose approximation property is preserved, but whose approximation constants cannot be uniformly bounded. We establish the required transfer principles, including the preservation of approximation bounds under complemented coordinate projections and the descent of the bounded approximation property from a dual space to its predual. We also provide an explicit complex-scalar version of the construction. The argument separates the external quantitative renorming input from the direct-sum construction and its functional-analytic consequences. In particular, the resulting example demonstrates that the approximation property of a dual space does not imply its bounded approximation property, even when the predual is separable and both spaces possess the approximation property.

fields
Functional Analysis
pdf ↗ page arXiv:2611.07930 ↗

We prove the length conjecture of J. Chang, T. Lyons and H. Ni for paths of bounded variation for arbitrary real Banach spaces and any reasonable tensor (algebra) norms. More precisely, we establish the missing lower bound and in this way show that the length of a tree-reduced path is recovered from the asymptotic behaviour of its normalised signature. The main technical argument uses the norm duality and shuffle algebra to express natural powers of polygonal length approximations as combinations of linear functionals acting on restricted signatures, which are then glued by suitable integrations into a linear functional of controlled growth acting on the full signature. In particular, the full proof combines tools of classical analysis and functional analysis, such as localisation by continuous cutoff functions, Stone-Weierstrass approximation of continuous functions by signature polynomials, the Hahn-Banach theorem and a support detection lemma for compactly supported real signed Borel measures.

comments
31 pages
fields
Functional Analysis, Classical Analysis and ODEs
pdf ↗ page arXiv:2611.08030 ↗

A number of results is presented concerning the extreme behaviour of products of linear operators or relations from one Hilbert space ${\mathfrak H}$ to another Hilbert space ${\mathfrak K}$. It is shown that there exists a relation $T$ from ${\mathfrak H}$ to ${\mathfrak K}$ such that the closure of the nonnegative relation $T^*T$ in ${\mathfrak H}$ is not equal to the nonnegative selfadjoint relation $T^*T^{**}$; moreover, this may happen for a singular but also for a regular (i.e., closable) relation $T$. In fact, as an extreme situation the paper contains a construction of a densely defined closable operator $T$, such that ${\rm dom\,}T^*T=\{0\}$, while $T^*T^{**}$ is a densely defined selfadjoint operator. Likewise it is shown that there exist a linear relation $T$ from ${\mathfrak H}$ to ${\mathfrak K}$ and a nonnegative everywhere defined contraction $R$ in ${\mathfrak K}$ for which the product $RT$ is singular and the product $R^{1/2} T$ is regular. Such results illustrate situations that can arise in the Lebesgue type decomposition of linear relations and the Lebesgue type decomposition of semibounded forms.

comments
13 pages
fields
Functional Analysis
pdf ↗ page arXiv:2611.08225 ↗

We study dimensional concavity properties of even log-concave measures within the framework of weighted centroaffine differential geometry. We express the relevant local concavity inequalities in terms of a weighted Hilbert–Brunn–Minkowski operator and establish an anisotropic Poincaré inequality for odd functions and a strengthened Brascamp–Lieb inequality for even functions. As applications, we obtain quantitative dimensional concavity estimates for a class of geometric measures and measures with pinched Hessian, extending results of Cordero-Erausquin and Rotem for rotationally invariant measures and Eskenazis and Moschidis for the Gaussian measure.

comments
Comments are welcome!
fields
Functional Analysis
pdf ↗ page arXiv:2611.08283 ↗

In this paper unbounded representations of quasi *-algebras on an indefinite inner product space are introduced and studied. After introducing the partial *-algebra of unbounded closable operators sharing a common dense domain in a Krein space, the appropriate notion of *-representation in this setting is developed, and a GNS-type construction for quasi *-algebras based on $α$-positive sesquilinear forms is established. Furthermore, the notion of abstract conditional expectation is generalized to quasi *-algebras and how \emph{admissible} abstract conditional expectations $P$ give rise to representable $P$-forms is shown. The relation between $α$-positive forms and $P$-forms is also clarified. Several examples are provided.

comments
28 pages
fields
Functional Analysis
pdf ↗ page arXiv:2611.08329 ↗

Using a family of bipartite graphs constructed by hanging identical rooted trees at every vertex of a highly symmetric graph, we show that every sufficiently large real number is the norm of an idempotent Schur multiplier on the bounded operators on $\ell^2$. This answers a question previously raised by the author.

fields
Functional Analysis, Combinatorics, Operator Algebras
pdf ↗ page arXiv:2611.08419 ↗

We prove Conjecture 1.4 of Prochno and Strzelecki [J. Approx. Theory 277 (2022), 105736] on Gelfand numbers of embeddings between finite-dimensional Schatten classes. For $0<p\le1$ and indices bounded away from the ambient dimension by a fixed proportion, we obtain the conjectured lower bound, uniformly in the target exponent $2\le q\le\infty$. Together with the known upper estimate obtained by Hinrichs, Prochno and Vybíral [Math. Ann., 380(3-4):1563-1593, 2021], this gives the exact asymptotic order in the previously unresolved range. As a consequence, we obtain a complete all-index description by combining our result with the estimates of Hinrichs, Prochno and Vybíral and of Prochno and Strzelecki. The proof is deterministic and is based on Hilbert–Schmidt projections on rectangular matrix spaces.

comments
11 page, 1 figure
fields
Functional Analysis
pdf ↗ page arXiv:2611.08448 ↗

Representation Theory8

We summarize a coordinate-free classification of five-dimensional real Lie algebras up to isomorphism. Rather than solving the overdetermined quadratic Jacobi system modulo $\mathrm{GL}(5,\mathbb{R})$, the method uses Levi–Malcev theory and nilradical stratification. In the non-solvable case, the Levi factor is three-dimensional, $\mathfrak{so}(3)$ or $\mathfrak{sl}(2,\mathbb{R})$, acting on a two-dimensional radical; this yields exactly five classes, including the indecomposable $\mathfrak{sl}(2,\mathbb{R})\ltimes\mathbb{R}^2$. In the solvable case, Schur's bound on commutative subalgebras of $\mathfrak{gl}(2,\mathbb{R})$ implies $\dim\mathfrak{n}\in\{3,4,5\}$, eliminating one- and two-dimensional nilradicals. The nilpotent case $\dim\mathfrak{n}=5$ has nine classes: three decomposable and six indecomposable, including $\mathfrak{h}_5$ and two filiform algebras. For $\dim\mathfrak{n}=4$, classification reduces to normal forms of outer derivations for $\mathbb{R}^4$, $\mathfrak{h}_3\oplus\mathbb{R}$, and $\mathfrak{n}_4$ under automorphism and scaling. For $\dim\mathfrak{n}=3$, it uses two-dimensional abelian subalgebras of $\mathrm{OutDer}(\mathfrak{n})$ containing no nonzero nilpotent operators. The work also computes minimal faithful representation dimensions $μ(\mathfrak{g})$, gives a dictionary with classical tables, and designs a deterministic identification algorithm.

fields
Representation Theory, Differential Geometry
pdf ↗ page arXiv:2611.06971 ↗

We compute higher theta lifts of ladder representations of general linear groups over division algebras over non-archimedean local fields. We achieve this by constructing BGG-type coresolutions and proving that their terms are acyclic for the covariant theta functor. We then compute the resulting complex explicitly, determining the higher theta lifts and the degrees in which they are nonzero.

fields
Representation Theory
pdf ↗ page arXiv:2611.07090 ↗

We prove that the action of the cactus group on a Coxeter group defined by Losev and Bonnafé can be realised as an action on the asymptotic algebra of the Hecke algebra, thus resolving a conjecture of Rouquier and White. We prove this in the setting of general (potentially unequal) parameters, conditional on Lusztig's conjectures P1-P15, which are known to hold in the equal parameter case.

fields
Representation Theory
pdf ↗ page arXiv:2611.07439 ↗

Let $K=k((t))$, where $k$ is algebraically closed of characteristic zero, and let $Λ$ be an acyclic ramified species over $K$. An indecomposable $Λ$-module can split after extending $K$. The number of its distinct geometric indecomposable summands is the degree of its residue endomorphism field $\operatorname{End}_Λ(M)/\operatorname{rad}\operatorname{End}_Λ(M)$. We use the positive roots and cyclic automorphism of the unfolded quiver to determine the number of isomorphism classes at every nonzero rank and residue degree $e\ge1$: it is zero, an explicitly computable finite number, or $|K|$. For every invariant imaginary root of the unfolding, including at nonprimitive ranks, we construct a smooth invariant family of absolutely indecomposable representations whose rational quotient by the change-of-bases group has dimension $1-q_d(β)$. We characterize ranks admitting absolutely Schur representations by the Schur roots of the unfolded quiver. For the associated species over finite fields, we realize the polynomials counting absolutely indecomposable representations as graded traces of the unfolding automorphism on preprojective BPS cohomology. Galois-orbit formulas and Möbius inversion then recover the counts at each residue degree. Finally, we prove an all-rank symmetrizable analogue of Kac's constant-term conjecture: the constant term of the polynomial counting all indecomposables, rather than only absolutely indecomposable ones, equals the corresponding Kac–Moody root multiplicity. This constant is independent of the symmetrizer and modulation.

comments
47 pages, comments are welcome
fields
Representation Theory, Algebraic Geometry, Rings and Algebras
pdf ↗ page arXiv:2611.07549 ↗

Queffelec–Sartori and Tubbenhauer–Vaz–Wedrich give a diagrammatic web category via generators and relations to describe the subcategory of Rep($\mathfrak{gl}(m|n)$) generated by tensor products of exterior powers of the defining representation. We construct the first explicit bases for homomorphism spaces in the web category for $\mathfrak{gl}(1|1)$, giving an alternative diagrammatic presentation of this web category, and explore connections to the combinatorics of planar networks. Our basis comes equipped with an algorithm to decompose any web into a linear combination of basis webs, giving an effective way to compute in the $\mathfrak{gl}(1|1)$-skein theory.

comments
17 pages
fields
Representation Theory, Combinatorics
pdf ↗ page arXiv:2611.07664 ↗

Let $\mathfrak{g}$ be a finite-dimensional simple Lie algebra of type $B$ or $C$ over ${\mathbb C}$. Given $κ\in{\mathbb C}$, let $V^κ(\mathfrak{g})$ be the universal affine vertex algebra associated with $\mathfrak{g}$ at level $κ$. We characterize the weights of minimal singular vectors of $V^κ(\mathfrak{g})$ with respect to the generalized Cartan matrix, when $V^κ(\mathfrak{g})$ is not simple and $κ$ is not critical. Furthermore, we determine the corresponding longest Weyl elements in the Kashiwara-Tanisaki character theorem of the singular vectors.

fields
Representation Theory, Quantum Algebra
pdf ↗ page arXiv:2611.07917 ↗

For projective morphisms $π:Y\to X=\operatorname{Spec}R$ over noetherian complete local rings, with fibres of dimension at most one and $\mathbf{R}π_*\mathcal{O}_Y\cong\mathcal{O}_X$, we construct a tilting fan from cones of pretilting bundles. Maximal cones parametrize tilting bundles, and facet adjacency encodes irreducible mutation. Completeness in positive rank is equivalent to finiteness of tilting bundles modulo line-bundle twists. For closed fibres that are split ADE highest-root cycles or their Stein contractions, we identify tilting bundles with chambers of restricted affine Tits cones and determine the fan. Over normal Gorenstein complete local domains, reflexive pushforward takes pretilting bundles on normal projective birational models with fibres of dimension at most one and $π^!\mathcal{O}_X\cong\mathcal{O}_Y$ to modifying modules. Pushforwards of canonical tilting bundles recover models as module blowups and detect contractions by additive containment, yielding an Auslander–McKay embedding. The hypersurfaces $R=S[[u,v]]/(uv-f)$, with $S$ a regular complete local ring and $f\in S$ a nonzero nonunit, generalize type $A$ Du Val surfaces and $cA_n$ threefolds to arbitrary dimension. We classify modifying modules by divisor flags and rank-one reflexive twists; flag blowups turn the embedding into an equivalence. If $f$ has $r\ge2$ prime factors counted with multiplicity, reflexive pushforward on any maximal-flag model gives a mutation-preserving surjection from tilting bundles, parametrized by chambers of affine type $\widetilde{A}_{r-1}$, to maximal modifying modules. Its fibres are affine Weyl group orbits determined by repeated factors. All maximal-flag models and maximal modification algebras are derived equivalent. These algebras are noncommutative crepant resolutions exactly when $S/(p)$ is regular for every prime factor $p$ of $f$.

comments
65 pages
fields
Representation Theory, Commutative Algebra, Algebraic Geometry, Rings and Algebras
pdf ↗ page arXiv:2611.07987 ↗

It was shown by Braverman-Finkelberg-Nakajima that the regular perverse sheaf on the affine Grassmannian of type A is isomorphic to the ring object associated to a Gelfand-Tsetlin quiver. We use this result to identify, for any finite dimensional representation $\mathbf{N}$ of a product of general linear groups, the relative Langlands dual $(T^*\mathbf{N})^\vee$ of $T^*\mathbf{N}$ with the Coulomb branch of a suitable representation. By a result of Bellamy, this implies that $(T^*\mathbf{N})^\vee$ has symplectic singularities.

comments
4 pages
fields
Representation Theory
pdf ↗ page arXiv:2611.08464 ↗

Classical Analysis and ODEs7

We prove a lower bound of $Ω_d((\log N)^{d-1/2})$ for the worst-case combinatorial discrepancy of $N$-point sets with respect to anchored boxes in every fixed dimension $d\ge3$. Together with Nikolov's upper bound of the same order and Wei's planar lower bound, this settles Tusnády's problem: the discrepancy of axis-parallel boxes is $Θ_d((\log N)^{d-1/2})$ in every fixed dimension $d\ge2$. The proof combines smoothed product test functions, sampling discretization, and an entropy argument for conditioned Gaussian vectors.

comments
33 pages
fields
Classical Analysis and ODEs, Combinatorics
pdf ↗ page arXiv:2611.06998 ↗

We demonstrate that the square function estimate for the light cone in $\R^d$, $d\ge3$, at the exponent $p=2(d-1)/(d-2)$, with loss $[\log(2R)]^{5/p}$ can be deduced from local-smoothing and auxiliary-energy statements in the October 25, 2026 OpenAI manuscript. Exact affine sections preserve these logarithmic bounds and give corresponding estimates for spheres, ellipsoids, elliptic paraboloids, and compact patches of two-sheet elliptic hyperboloids in $\R^{d-1}$.

fields
Classical Analysis and ODEs
pdf ↗ page arXiv:2611.07085 ↗

We give a geometric proof of the radial Kunze–Stein phenomenon on the good vertices of a thick, locally finite, regular affine building of irreducible type. The proof uses a positive spherical function, its joint eigenvalues for sphere averages, and the growth of vector-distance spheres. Conjugation by this function and complex interpolation yield a weighted $\ell^2$ estimate and the boundedness of radial operators with $\ell^p$ kernels for $1\leq p<2$. We also present proofs by the averaging operators norms, by the Schur test and by Helgason transform. A further interpolation gives weighted mixed-exponent estimates. No transitivity assumption on the automorphism group is required; in particular the argument applies to buildings of type $\widetilde A_2$ without a Bruhat–Tits realization.

comments
14 pages, submitted to Colloquium Mathematicum
fields
Classical Analysis and ODEs
pdf ↗ page arXiv:2611.07889 ↗

In a previous paper we characterized the lower-upper (LU) and upper-lower (UL) factorizations of the infinitesimal generator of a possibly nonconservative birth-death process whose Darboux transformations are again generators of birth-death processes. Here we apply these results to several classical models with linear rates and explicitly known spectral measures: the $M/M/\infty$ queue (Charlier polynomials), the linear birth-death processes (Meixner and Laguerre polynomials), and the linear models of Ismail, Letessier and Valent (associated Laguerre polynomials). For each of them we compute the factors in closed form, determine the admissible ranges of the parameters, and describe the rates, spectral measures and orthogonal polynomials of the Darboux-transformed processes.

comments
33 pages, 4 figures
fields
Classical Analysis and ODEs, Probability
pdf ↗ page arXiv:2611.07974 ↗

We investigate a broad class of linear inhomogeneous boundary-value problems of order $r$ for systems of ordinary differential equations depending on a parameter $μ$ belonging to an abstract metric space $\mathcal M$. The solutions belong to the spaces of continuously differentiable vector-valued functions $(C^{(n+r)})^m$, $n\in\mathbb{N}\cup\{0\}$, $m, r \in \mathbb{N}$. The boundary conditions have the most general form $By=c$, where $B$ is an arbitrary continuous linear operator from $(C^{(n+r)})^m$ to $\mathbb{C}^{rm}$. Thus, they may contain derivatives of the unknown vector-valued function of integer and/or fractional orders up to $n+r$ inclusive. We establish necessary and sufficient conditions for the continuous dependence of solutions on the parameter $μ$. Furthermore, we prove that the solutions of the original problems can be approximated in the space $(C^{(n+r)})^m$ by solutions of systems of ordinary differential equations with polynomial coefficients, polynomial right-hand sides, and multipoint boundary conditions, where the approximating problems are constructed independently of the right-hand sides of the original problem.

fields
Classical Analysis and ODEs
pdf ↗ page arXiv:2611.07991 ↗

We prove a pointwise domination of Poisson variances for Hörmander–Mikhlin multipliers on the real line. The estimate holds in operator order for operator-valued functions and gives a direct Fourier-side proof of BMO boundedness. The same argument gives an algebraic proof of the one-dimensional Schur multiplier estimate of Conde-Alonso, González-Pérez, Parcet and Tablate ( Annals of Mathematics, 2023), formulated here using Poisson BMO. The endpoint estimates are proved for $L^2$ functions and Hilbert–Schmidt operators; interpolation gives the corresponding $L^p$ and Schatten estimates.

fields
Classical Analysis and ODEs, Functional Analysis
pdf ↗ page arXiv:2611.08192 ↗

Building on our earlier work, we present an $L^2$-$L^1$ inequality for radial Fourier eigenfunctions, relating this to the sign uncertainty principle of Bourgain, Clozel and Kahane. We also propose a generalisation of the aforementioned uncertainty principle.

comments
23 pages, comments welcome
fields
Classical Analysis and ODEs
pdf ↗ page arXiv:2611.08224 ↗

Logic7

This paper proves asymptotically optimal linear degree lower bounds for Nullstellensatz and Polynomial Calculus, yielding a four-way polynomial-versus-exponential proof-size classification for uniform polynomial consequences over fields of positive characteristic. The size bounds use full, explicit coefficient lists, including zeros. Apart from a fixed finite initial range, exactly one of four cases occurs: $\mathrm T_1$, bounded degree; $\mathrm T_2$, degree at least linear in the universe size $n$; $\mathrm T_3$, constant degree on some residue classes and at least linear degree on the others; or $\mathrm T_4$, infinitely many recurring finite degrees, with bounded-degree subsequences and finite $Ω(n)$ peaks at arbitrarily large sizes. Thus the size patterns are polynomial bounds, exponential lower bounds, a periodic mixture, or polynomial subsequences with finite exponential peaks, respectively. Underivable targets have degree $\infty$, and both proof systems assign the same type. The result strengthens my 2008–2009 refutation classification and extends it to uniform consequences. The proof uses degree-preserving lifting with onset and period bounds linear in the degree. A common $p$-adic hardness spectrum describes the four types topologically. Distinguishing any two types is undecidable already for refutations. The linear degree bounds, four-way classification and pairwise undecidability results have been formalised and verified in Lean. The Lean formalisation, verification records and an accompanying tutorial are available in the supporting materials (https://doi.org/10.5281/zenodo.23965838).

comments
78 pages. Tutorial, Lean formalisation and verification records: https://doi.org/10.5281/zenodo.23965838
fields
Logic
pdf ↗ page arXiv:2611.07280 ↗

A $k$-hypertournament is a higher-arity generalisation of a tournament, consisting of a set together with an irreflexive $k$-ary relation such that the automorphism group of each substructure of size $k$ is the alternating group $\mathrm{A}_k$. We show that, for each $k \geq 3$, there are continuum-many non-isomorphic countable ultrahomogeneous $k$-hypertournaments, answering a question of Cherlin.

fields
Logic, Combinatorics
pdf ↗ page arXiv:2611.07292 ↗

We give a true $Π^0_1$ sentence $\varphi$ which is independent of Peano Arithmetic. Moreover, if $\widehatφ$ denotes its interpretation on $ω$ as a sentence of set theory, then $\widehatφ$ is a $Π^0_1$ sentence independent of ZFC.

fields
Logic
pdf ↗ page arXiv:2611.07321 ↗

${T}_n$-Heyting algebras form a subvariety of Heyting algebras and are characterized by the property that a chain of prime filters in such an algebra has at most $n-1$ elements. We study some basic properties of these algebras and determine the structure and number of elements of the free ${T}_n$-Heyting algebra with a finite number $r$ of free generators, $FT_n(r)$.

comments
32 pages, 4 tables, 8 figures
fields
Logic
pdf ↗ page arXiv:2611.07915 ↗

We prove equivalences of asymmetric partition relations involving natural numbers and products of weakly compact cardinals $κ$, infinite cardinals $λ<κ$ and natural numbers to certain classes of finitary problems in the theory of edge-coloured digraphs. We are able to determine three classes of Ramsey numbers exactly, typical examples in these classes are $r(ω^2 2, 3)$, $r(κλ2, 3)$ and $r(κλ3, 3)$. We moreover provide general upper bounds for $r(ω^2 m, 3)$ and $r(κλm, n)$.

comments
36 pages, 8 figures, 2 tables
journal
Weinert, T. Idiosynchromatic Poetry. Combinatorica 34, 707--742 (2014)
fields
Logic, Combinatorics
pdf ↗ page arXiv:2611.08079 ↗

Using techniques from model theory, we prove analogues of Rado's and Sárközy's theorems on solutions to polynomial equations in dense sets of definable groups and fields equipped with well-behaved measures. As a special case, we adapt our results to pseudofinite structures, to deduce non-qualitative asymptotic results.

comments
15 pages
fields
Logic
pdf ↗ page arXiv:2611.08211 ↗

The Arthur-Nimue-Merlin degrees are a generalization of the Turing degrees introduced by Kihara as a tangible description of the partially ordered set of Lawvere-Tierney topologies on the effective topos (equivalently, subtoposes of the effective topos). They are defined in terms of a three-player game that introduces both angelic and demonic non-determinism into oracle queries. We construct an order embedding of the Turing degrees with their order reversed into the Arthur-Nimue-Merlin degrees, whose image we call the "co-Turing degrees"; we then study the order relationship of these co-Turing degrees with the (naturally embedded) Turing degrees within the Arthur-Nimue-Merlin degrees. This is a short version of arXiv:2603.19946, focused on the main results.

comments
18 pages, comments welcome. Conference version of arXiv:2603.19946, accepted at CSL 2027
fields
Logic
pdf ↗ page arXiv:2611.08319 ↗

Algebraic Topology6

Data arising in scientific applications often exhibits complex nonlinear structure. Identifying and parameterizing said nonlinearities can enable better modeling, enhance predictive analysis, data interpolation, denoising, and anomaly detection. In this paper, we develop theoretical and computational methods to detect and parametrize Euclidean point cloud data lying near a wedge sum of tori. Specifically, we use persistent cohomology and cup products to identify the distinct toroidal components, and on each component, we optimize over low-energy multivariate trigonometric polynomials for best data fit. The ensuing optimization problem reduces to generalized ridge regression, which can be solved efficiently via standard techniques. We demonstrate the utility of the proposed framework in both synthetic and real datasets, including interpolation of low frame-rate recurrent videos, attractor recovery in quasiperiodic dynamical systems, and the parametrization of gene-space cycle data for parasites and humans.

fields
Algebraic Topology
pdf ↗ page arXiv:2611.06990 ↗

We study the topology of the normalized space of real nonnegative forms of degree $2d$ in $n$ variables that are not sums of squares. Whenever it is nonempty, we prove that it is path-connected but not acyclic, and in particular not contractible. We give explicit bounds on the degrees in which its reduced homology can be nonzero. In the two minimal nonempty cases, we show an embedding of an exterior algebra into the cohomology ring, and prove containment of the binary octahedral and binary icosahedral groups, respectively, in its fundamental group.

comments
23 pages
fields
Algebraic Topology, Algebraic Geometry
pdf ↗ page arXiv:2611.07252 ↗

We explicitly determine the complex bordism class of the $n$-th symmetric product of a complex curve of genus $g$. In particular, we provide an explicit formula expressing this class as an integral linear combination of bordism classes of products of complex projective spaces. Similarly, we determine the unoriented bordism class of the $n$-th symmetric product of a non-orientable surface of genus $g$ (i.e., $\#_{i=1}^{g+1}\mathbb R P^2$). It turns out that these classes are also expressed as a $\mathbb Z_2$-linear combination of bordism classes of products of even-dimensional real projective spaces. We also give an affirmative answer to a non-orientable version of the Blagojević–Grujić–Živaljević conjecture stating that the topological type of any symmetric product of a noncompact surface of finite type determines the topological type of the surface itself.

fields
Algebraic Topology
pdf ↗ page arXiv:2611.07813 ↗

In this paper, we establish a topological classification of Bazaikin spaces whose sixth cohomology groups have orders prime to \(5\). Moreover, we obtain an estimate for the number of diffeomorphism types among these spaces.

comments
This is a revised version of our earlier manuscript, arXiv:2307.13303. A gap in the original proof of the earlier version has been filled
fields
Algebraic Topology, Differential Geometry, Geometric Topology
pdf ↗ page arXiv:2611.07898 ↗

This is the first part of a series of expository notes on homotopical and categorical algebra. It charts a course from ordinary algebraic structures such as monoids, groups, and rings to their various homotopy-coherent generalizations and equivariant versions. The emphasis is on the ideas that lead to and connect the constructions, as well as where and how they appear in nature.

comments
84 pages. Comments welcome
fields
Algebraic Topology, Category Theory
pdf ↗ page arXiv:2611.07927 ↗

We study smooth structures on the topological manifold underlying complex projective space. We conjecture that every complex projective space $\mathbb{CP}^n$ with $n>3$ admits an exotic smooth structure. We prove two results providing strong evidence for this conjecture. First, the set of complex dimensions $n$ for which $\mathbb{CP}^n$ admits an exotic smooth structure has natural density one. Second, we explicitly identify $266$ congruence classes modulo $288$ such that $\mathbb{CP}^n$ admits an exotic smooth structure whenever $n>3$ lies in one of these classes. These classes cover more than $92\%$ of complex dimensions, improving the previously known lower bound of approximately $36\%$. We prove these results by first relating the concordance group of $\mathbb{CP}^n$ to suitable Atiyah–Hirzebruch filtration subgroups of its stable cohomotopy group, and then applying computations in chromatic homotopy theory at height $2$.

fields
Algebraic Topology
pdf ↗ page arXiv:2611.08406 ↗

Group Theory6

In this note, we determine the finite abelian groups $G$ such that for every subgroup $H\leq G$ there exists an automorphism $f\in{\rm Aut}(G)$ with ${\rm Fix}(f)=H$. This provides a partial answer to a problem posed by Deaconescu and Walls. Two generalizations of this problem are also proposed.

fields
Group Theory
pdf ↗ page arXiv:2611.07224 ↗

Let F be a saturated fusion system on a finite 2-group S, and let B(F) be the Burnside ring of F. Barsotti and Carman proved that when S is abelian, B(F)^*=B(S)^* if and only if F=F_S(S). Barsotti and Carman asked on page 728 of their paper whether there exists a counterexample to Corollary 9.3 for a nonabelian 2-group. We have given a positive answer. Our main conclusion is that the following are equivalent for all finite 2-groups: (1)F=F_S(S); (2)B(F)^*=B(S)^*; (3)For every maximal subgroup M of S, 1-[S/M] is in B(F)^*.

fields
Group Theory
pdf ↗ page arXiv:2611.07970 ↗

We prove that every countably infinite group endowed with a topology of countable weight in which every nonempty open set is infinite admits an exact factorization into two dense infinite subsets. More generally, every finite partial exact factorization with nonempty factors extends to an exact factorization of the whole group in which both factors meet every member of a prescribed at most countable family of infinite subsets in infinitely many points. This settles the countable case of a question of Protasov and Slobodianiuk. The proof uses the finite graph Ramsey theorem, invariant finitely additive probability measures on amenable groups, and a structural lemma on quadratic coverings based on Neumann's coset covering lemma.

comments
9 pages
fields
Group Theory
pdf ↗ page arXiv:2611.08171 ↗

We give a human-readable exposition of OpenAI's construction of a group algebra in characteristic two that is not directly finite. We then adapt the argument to show that, for every prime $p$, there exists a hyperbolic group $G_p$ such that $\mathbb{F}_p[G_p]$ is not directly finite.

comments
32 pages
fields
Group Theory, Rings and Algebras
pdf ↗ page arXiv:2611.08268 ↗

We prove that for a Chevalley–Demazure group scheme $\mathcal{G}$ of type $G_2$, $F_4$ or $E_8$ and a Euclidean ring $\mathcal{O}$ with field of fractions $K$, the Steinberg module of $\mathcal{G}(K)$ is generated by integral apartment classes. As a consequence, the top-dimensional rational cohomology of $\mathcal{G}(\mathcal{O})$ vanishes if $\mathcal{O}$ is a ring of integers, extending results of Tóth and of Brück–Santos Rego–Sroka.

fields
Group Theory, Number Theory
pdf ↗ page arXiv:2611.08381 ↗

Let $G$ be a not necessary full automorphism group of a compact Riemann surface of genus $g$. We found sharp bounds for $|G|$ in the terms of $g$ when $G$ belongs to one of the following classes of finite groups sharing similar properties with supersoluble groups: 2-nilpotent, with the Sylow tower of supersoluble type, with $\mathbb{P}$-subnormal cyclic primary subgroups, widely supersoluble, with submodular Sylow subgroups, with submodular cyclic primary subgroups and strongly supersoluble.

fields
Group Theory, Algebraic Geometry
pdf ↗ page arXiv:2611.08391 ↗

Metric Geometry6

We provide sharp estimates on the number of vertices needed to approximate the Euclidean unit ball $B_d\subset\mathbb R^d$ with relative symmetric-difference volume error at most $\varepsilon\in(0,1)$. We show that the minimum vertex budget $N_\varepsilon(d)$ satisfies $$\log N_\varepsilon(d)\asymp d\log\left(1+\frac1{\varepsilon\sqrt d}\right).$$ In particular, $\exp(O(\sqrt d/\varepsilon))$ vertices suffice for every $\varepsilon\in(0,1)$ and all sufficiently large $d$. We also prove that achieving a fixed error $\varepsilon\in(0,1)$ requires at least $Ω\bigl((\tfrac{1+\varepsilon}{2\sqrt\varepsilon})^{d-1}\bigr)$ facets, an exponential number in $d$. We extend our vertex approximation result to Gaussian symmetric-difference error, and give a positive answer to the question of De, Nadimpalli, and Servedio about subexponential vertex complexity for approximating Euclidean balls. We also disprove a Macbeath-type extremality principle for unrestricted vertex approximation. Whereas the ball uniformly maximizes the inscribed approximation error among all convex bodies (Macbeath, 1951), we show that approximating the hypercube $[-1,1]^d$ to relative error $\varepsilon$, with $δ:=1-\varepsilon$, requires $N\ge\fracδ{2}\exp(Ω(δ^2d/\log^2(2/δ)))$ vertices whenever $d\ge16\log(2/δ)$. Thus, when $d$ is large, the cube is harder to approximate than the ball in this sense. Finally, we determine the sharp dimensional order of the classical Laguerre-Delone triangulation number, showing that $\mathrm{ldel}_{d-1}=Θ(d^{-1/2})$ as $d\to\infty$. This improves both the upper bound of Ludwig, Schütt, and Werner and the lower bound of Böröczky by a factor of order $\sqrt d$. By classical asymptotic approximation theorems (Ludwig, 1999), this improvement applies to every convex body in $\mathbb R^d$ with $C^2$ boundary and positive Gaussian curvature.

comments
71 pages, 1 figure, 1 table. v2: exposition and references revised for a broader audience
fields
Metric Geometry, Probability
pdf ↗ page arXiv:2610.34651 ↗

We characterize neutral planes by three metric conditions: existence of geodesics, side-side-side congruence, and uniqueness of reflection.

comments
5 pages, 8 figures
fields
Metric Geometry, Differential Geometry
pdf ↗ page arXiv:2611.07099 ↗

We prove the existence of a sharp threshold $p_0 \approx 26.265\ldots$, such that for every $n\in\mathbb{N}$ and $p\geq p_0$, the volume maximizing section of the unit ball of $\ell_p^n$ is the section with Ball's hyperplane $(\frac{e_1+e_2}{\sqrt{2}})^\perp$, whereas for $p\in(2,p_0)$ the asymptotic maximizer as the ambient dimension $n\to\infty$ is the hyperplane orthogonal to the diagonal vector $\frac{e_1+\cdots+e_n}{\sqrt{n}}$.

fields
Metric Geometry, Functional Analysis
pdf ↗ page arXiv:2611.07619 ↗

Rigid-origami folding has versatile applications in the fields of architecture, engineering, medicine and aerospace. Given the design parameters of a single waterbomb cell, the global geometry of the origami tessellated shell at an intermediate state of folding is directly determined. Even though existing numerical methods of rigid-origami can be used to describe geometric forms, algebraic solutions of the waterbomb folding kinematics and tessellations reduce the algorithmic complexity and determine the range of admissible configurations exactly. The existing analytical solution for a symmetrically folded waterbomb cell and tessellation is extended to the case of a cell folded semi-symmetrically, i.e. a cell with a 180° rotation symmetry. The algebraic derivation provides the solution for the entire folding process. The resulting functional relationship between the design parameters and the 3D cell geometry of the folded waterbomb cell enables efficient studies of achievable shell geometries. This development facilitates a deeper understanding of the correspondence between the geometric parameters of a single waterbomb cell and the statically relevant geometric properties, i.e. the shell curvature and the effective cross-sectional height. The systematic description of this correspondence provides the basis for innovative design concepts using modern cementitious composites, e.g. textile-reinforced concrete, to produce thin-walled, high-performance, lightweight shells.

comments
Accepted for publication in Proceedings of the Royal Society A. 24 pages. Author accepted manuscript
fields
Metric Geometry
pdf ↗ page arXiv:2611.08050 ↗

We construct a compact set of diameter $1$ in $\mathbb{R}^5$ which cannot be covered by six Euclidean balls of diameter $1$. Its convex hull is a full-dimensional convex body satisfying the same property. Moreover, a finite subset of diameter $1$ cannot be covered by six closed balls even when their radius is increased to $0.50001$. The compact set is the quadratic image of the unit sphere in the space of symmetric $3\times3$ matrices of trace zero. A hypothetical cover leads to twelve spherical ellipses covering ${\mathbb S}^2$ and arranged in six antipodal pairs. Using Euler's formula, we replace them by twelve pentagonal regions contained in the ellipses. An area estimate then gives a contradiction.

comments
14 pages
fields
Metric Geometry
pdf ↗ page arXiv:2611.08198 ↗

In this paper, we continue our investigation of the structure of Thurston geometries with non-constant curvature. We generalize the notion of cylinders, prove that they contain the types of geodesic curves. Moreover, we formulate and prove a statement analogous to Pappus's hexagon theorem for $\SLR$ geometry.

comments
17 pages, 4 figures
fields
Metric Geometry
pdf ↗ page arXiv:2611.08234 ↗

Statistics Theory6

We develop specification tests for parametric conditional tail index models of heavy-tailed distributions. The tests are based on a projected empirical process that removes the first-order effect of estimating the null parameter. Within the conditional Hall class, we identify conditions under which projection also eliminates the leading Pareto approximation bias in the test process, permitting lower thresholds while preserving asymptotically correct size. We establish the validity of the tests via multiplier bootstrap. For log-linear models, bootstrap validity extends to settings in which covariates among observations above a common threshold concentrate and the smallest eigenvalue of the information matrix tends to zero. We further propose a data-driven threshold selection rule based on an estimate of the Pareto approximation bias after projection. The rule allows lower thresholds to retain more observations for testing while limiting size distortions, with potential gains in power. The finite sample performance of the tests is examined through simulations and a real data application.

fields
Statistics Theory
pdf ↗ page arXiv:2611.07219 ↗

Testing seeks to distinguish distributions, whereas matching seeks to identify their common mass. I show that this apparent opposition provides a direct proof of the Neyman–Pearson lemma. Maximal partial matching between the alternative distribution and a scaled null distribution separates them into a common component and two mutually singular residuals. An elementary dual problem then turns this separation into a testing rule: complementary slackness forces rejection on the alternative residual and non-rejection on the scaled-null residual, leaving freedom to randomize at the likelihood-ratio threshold. Thus the optimal test emerges from maximal matching rather than being proposed in advance. The matching profile also determines the optimal power envelope.

fields
Statistics Theory
pdf ↗ page arXiv:2611.07368 ↗

We study the cost of certifying a scalar target driven by a product of centered Bernoulli means when an unobserved terminal gate leaves irreducible uncertainty. Each factor can be queried independently at a price. We derive the exact identified interval and determine when a requested accuracy is feasible. At every fixed strictly feasible interior instance, the optimal high-confidence expected cost is $(1+o(1))\log(1/δ)/D^*$, where $D^*$ is an answer-and-allocation information game; a globally honest procedure attains this benchmark. At product cancellation, we solve the game through a one-dimensional inverse-KL equation and explicit cost weights. For two tickets, quartic projections and rational bounds give checkable global optimality gaps. Joint confidence–precision limits expose a failure of the analyzed fixed pilot: it can cost asymptotically at least twice the optimum. A charged, precision-aware pilot with checkpoint certification restores optimal expected cost in exact arithmetic when $\log\{e+\log(1/g)\}=o(\log(1/δ))$ for effective precision $g$, including exponential and certain double-exponential precision scales. Pooled rescue preserves validity and cannot increase observation cost on coupled paths. Implemented auditors verify returns by rational inequalities; reproducible experiments show both savings and finite-confidence overhead. These guarantees concern the specified independent-reference model and its scalar target.

fields
Statistics Theory
pdf ↗ page arXiv:2611.07567 ↗

We record two observations concerning \cite{HLW24}. First, the proof of its Theorem~1 yields the population linearization (13)--(14) at \textit{any critical point} of the Fréchet function $F_μ$, under its remaining geometric and measure-theoretic assumptions; neither isolation nor a nonsingular Hessian of $F_μ$ is needed. Second, its Theorem~2 is given a shorter proof by applying the M-estimator theorem \cite[Theorem~5.23, pp.~53–54]{vdV98} to negative squared distance in normal coordinates, as opposed to the Donsker argument used there. We verify the pointwise differentiability, local square-integrable Lipschitz bound, second order expansion of population loss, consistency, and empirical maximization hypotheses. The resulting covariance retains the cut-locus correction. The same argument applies to suitably selected local sample Fréchet minima when the population Hessian is positive definite.

fields
Statistics Theory
pdf ↗ page arXiv:2611.07956 ↗

Spearman's rho treats all ranks alike, yet in many comparisons of two rankings an agreement at the top matters more than one at the bottom. Blest's rank correlation builds this preference into Spearman's rho, and its symmetrized version $η$, introduced by Genest and Plante, does so without favoring either of the two rankings. How far can such a top-weighted coefficient stray from other rank correlations? We answer this question for Spearman's rho, Blomqvist's beta and Chatterjee's rank correlation $ξ$ by determining the exact set of values that $η$ and each of these coefficients can attain jointly. Given Spearman's rho, $η$ can differ from it by at most $27/128$, and the extreme values of $η$ are attained only when the two rankings agree perfectly on the top ranks and are completely reversed on the bottom ranks, or vice versa. Given Blomqvist's beta, $η$ obeys exactly the same sharp bounds as Spearman's rho. For Chatterjee's rank correlation, both boundary curves are explicit and are attained by copulas with densities. Here symmetrization has a surprising effect. Compared with Blest's original coefficient, $η$ has a narrower range at fixed Spearman's rho, but near independence a wider range at fixed Chatterjee's rank correlation, where $|η|$ can come close to $\sqrt{11ξ/5}$ while Blest's coefficient stays below $\sqrt{32ξ/15}$. The proofs combine explicit Kantorovich potentials from optimal transport with a convex relaxation that turns out to be exact.

comments
24 pages, 4 figures
fields
Statistics Theory, Probability
pdf ↗ page arXiv:2611.08026 ↗

Let $\mathbf X_1,\ldots,\mathbf X_n$ be independent random vectors in $\mathbb R^d$ with common mean $\boldsymbolμ_n$, where $d\ge2$ is a fixed integer. Let $\bar{\mathbf X}_n=n^{-1}\sum_{i=1}^n\mathbf X_i$ denote the sample mean and let \[ \widehatΣ_n=\frac1n\sum_{i=1}^n (\mathbf X_i-\bar{\mathbf X}_n) (\mathbf X_i-\bar{\mathbf X}_n)^\top \] denote the sample covariance matrix. Assume that the average covariance matrix is positive definite and that $\widehatΣ_n$ is nonsingular almost surely. For testing $H_0:\boldsymbolμ_n=\boldsymbolμ_0$, the Hotelling statistic is \[ T^2_{H,n} =n(\bar{\mathbf X}_n-\boldsymbolμ_0)^\top \widehatΣ_n^{-1} (\bar{\mathbf X}_n-\boldsymbolμ_0). \] Under $H_0$, we establish finite-sample Cramér-type moderate deviation results and uniform Berry–Esseen bounds under finite $p$th moments for $2<p\le3$. The moderate deviation error order is optimal over independent triangular arrays, as shown by explicit constructions. For identically distributed observations with a fixed distribution, we obtain \[ \sup_{0\le x\le b_n} \left| \frac{\mathbb P(T^2_{H,n}\ge x^2)} {\mathbb P(χ_d^2\ge x^2)}-1 \right| = O\!\left((1+b_n)^p n^{-(p-2)/2}\right) \] for every nonnegative sequence $b_n=o(n^{(p-2)/(2p)})$. The uniform Berry–Esseen bound is $O(n^{-(p-2)/2})$. The proof uses a radial Stein identity and a recursive error estimate. We apply the results to chi-squared calibration in simultaneous testing.

comments
56 pages
fields
Statistics Theory, Probability
pdf ↗ page arXiv:2611.08074 ↗

Category Theory5

Controlled subcircuits are commonplace in quantum algorithms as they represent the quantum analogue of branching statements (if then). We introduce a generalization of controlled subcircuits in the form of a binary operation which sends pairs of unitary quantum circuits to a generalized controlled operation, whose semantics are based on matrix exponentials and logarithms. The family of quantum circuits with generalized controlled operations and circuit exponentiation is denoted HQC for Hierarchical Quantum Circuits. We formalize HQC as a modified prop category using a novel category-theoretic construction, and provide a sound and complete equational theory for diagrammatic reasoning in HQC.

comments
In Proceedings QPL 2026, arXiv:2611.04069
journal
EPTCS 456, 2026, pp. 283-349
fields
Category Theory
pdf ↗ page arXiv:2611.07207 ↗

We construct an idempotent complete Frobenius exact category with an idempotent complete stable category in which a complete ideal cotorsion pair of object ideals induces a cotorsion pair of objects that fails to be complete. This gives a negative answer to the question of Fu, Guil Asensio, Herzog and Torrecillas under idempotent completeness. The construction uses a lemma on dense triangulated subcategories and an example of Bondarko and Sosnilo in which a weight structure fails to extend to the idempotent completion. The induced cotorsion pair is hereditary, and every object admits an ordinary precover from its left class and an ordinary preenvelope into its right class. We characterize which objects admit special precovers and which admit special preenvelopes.

comments
9 pages
fields
Category Theory
pdf ↗ page arXiv:2611.07505 ↗

We study self-similar spaces obtained by repeatedly gluing finitely many scaled copies of a pointed metric space along prescribed boundary points. A finite gluing specification defines endofunctors on suitable categories of pointed sets and pointed metric spaces. Under non-degeneracy and metric admissibility assumptions, the initial algebra of the set functor carries a canonical metric. Its Cauchy completion is a final coalgebra for the metric functor and, after forgetting the metric, also for the underlying set functor. We compare this construction with Iterated Function Systems. When the orbit algebra is injective and its comparison with the initial algebra is bi-Lipschitz, the attractor is isomorphic to the completed final coalgebra. This gives a categorical analogue of Hutchinson's theorem for overlaps captured by finitely many distinguished points. Examples include the unit interval, Cantor spaces, the Sierpiński gasket and $p$-adic integers.

comments
40 pages
fields
Category Theory
pdf ↗ page arXiv:2611.07584 ↗

Aguiar, Bergeron, and Sottile showed that the coalgebra of quasisymmetric functions is terminal among connected graded coalgebras with a zeta function. After developing the notions of gradings and zeta functions for decomposition spaces, we establish an objective homotopic version of this result. Namely, the decomposition space $Q$ of quasisymmetric functions is terminal: for any conservatively graded decomposition space $X$ with zeta function $F$, there is a unique graded ikeo-culf span $X \leftarrow W \rightarrow Q$ inducing $F$ from the special zeta function on $Q$. More generally, for gradings in any complete decomposition space $M$ with indecomposable identities, the terminal object is $J(M)$, the free decomposition space on the nondegenerate simplices of $M$; the quasisymmetric case is $Q=J(B\mathbb{N})$.

comments
44 pages. comments welcome!
fields
Category Theory, Algebraic Topology, Combinatorics
pdf ↗ page arXiv:2611.08063 ↗

Virtual ultracategories are a recent categorification of topological spaces seen as sets of points together with a convergence relation between points and ultrafilters. However, a topology can also be described in terms of convergence of filters: in this paper, we thus introduce an analogue of virtual ultracategories based on filters rather than ultrafilters. At the topological level, convergence of filters is completely determined by its restriction to ultrafilters. Our main result shows that, under the assumption of tautness, the same phenomenon holds at the categorified level, thereby offering a non-ultra-biased perspective on virtual ultracategories.

fields
Category Theory, General Topology
pdf ↗ page arXiv:2611.08124 ↗

Complex Variables5

We study hyperbolicity notions for compact complex manifolds defined by special Hermitian metrics and by holomorphic maps from $\mathbb C^{n-1}$ (divisorial hyperbolicity). For surfaces we show that divisorial and Brody hyperbolicity coincide and that Hermitian-symplectic and Kähler hyperbolicity coincide when $h^{2,0}=0$. In any dimension we prove obstructions to Gauduchon hyperbolicity, show that projective spaces are not divisorially hyperbolic, and give divisorially hyperbolic manifolds that are not balanced hyperbolic.

fields
Complex Variables
pdf ↗ page arXiv:2611.07098 ↗

Each homeomorphic parametrization of a Jordan curve via the unit circle $\mathbb{S}\subset \mathbb{S}^2$ extends to a homeomorphism of the entire $\mathbb{S}^2.$ We show that there exists a Jordan curve of positive area whose every parametrization via $\mathbb{S}$ admits a homeomorphic extension that has $p$-Sobolev regularity on $\mathbb{S}^2$ for all exponents $1\le p<2 .$ This is obtained via a new sharp extension result for the case of $p<2.$

comments
34 pages, 6 figures
fields
Complex Variables
pdf ↗ page arXiv:2611.07322 ↗

We disprove a conjecture of Ruscheweyh, Salinas and Sugawa concerning the universal convexity of the normalized generating function associated with the Ramanujan sequence $θ(n).$ More precisely, we prove that the function \[ σ(z)=\frac{1}{θ(1)-1/3}\sum_{n\geq1}(θ(n)-1/3)z^{n} \] is not universally convex on the slit domain $\mathbb{C}\backslash[1,\infty).$ The proof uses Watson\textquoteright s integral representation of $θ(n),$ asymptotic estimates for an auxiliary function $h,$ and the representation of derivatives of universally convex functions via logarithmic integrals. As a general extension of this result, we also show that the generalized Ramanujan generating function \[ σ_α(z)=\frac{1}{θ(1)-1/3}\sum_{n\geq1}(θ(n)-1/3)\frac{z^{n}}{n^α} \] is universally convex if and only if $α\geq1.$

comments
Comments are welcome
fields
Complex Variables
pdf ↗ page arXiv:2611.07479 ↗

We prove the Edholm–Mathew nonvanishing conjecture for the palindromic polynomials associated with rational Hartogs triangles. For relatively prime positive integers $m>n$, we show that $e^{-i(m-n)θ}Q_{m,n}(e^{iθ})$ is strictly positive for every real $θ$ and derive an explicit lower bound. The proof expresses this trigonometric polynomial as a finite average of products of unnormalized Fejér kernels. It follows that $Q_{m,n}$ has exactly $m-n$ zeros in the punctured unit disc, counted with multiplicity. As a consequence, a rational Hartogs triangle has a zero-free Bergman kernel if and only if its exponent is the reciprocal of a positive integer.

comments
15 pages
fields
Complex Variables
pdf ↗ page arXiv:2611.07480 ↗

We study the cross ratio of four independent points chosen uniformly on the Riemann sphere. A positive integral over a two-dimensional simplex gives its normalized planar density. We then evaluate this integral in terms of the Bloch–Wigner dilogarithm, providing a closed formula. We describe its symmetries, real-axis values, moment bounds and geometric interpretations. We use these to evaluate the mean ideal-tetrahedron volume directly from the density, recovering $π/6$, a recent theorem of Kabluchko and Schange. We then prove that the expected distance between two independently sampled hyperbolic geodesics in $\mathbb{H}^3$ is $4/3$, by evaluating the logarithmic potential of the cross-ratio distribution on the negative real axis. Further applications give the joint dihedral-angle density and an exact acute-tetrahedron probability, and a branched-cover probability law on the modular surface with exponentially small cusp tails.

fields
Complex Variables, Probability
pdf ↗ page arXiv:2611.07670 ↗

Symplectic Geometry5

Local mirror symmetry, originated from the work of Chiang-Klemm-Yau-Zaslow arXiv:9903053, indicated a potential relation between mirrors to a vector bundle and mirrors to its zero section. In this paper, we constructed a homological mirror to these local varieties and proved the corresponding Homological Mirror Symmetry statement. As an application, we constructed infinite families of exotic Weinstein sectors as mirrors to Hirzebruch surfaces.

comments
56 pages, 5 figures, comment welcome!
fields
Symplectic Geometry, Algebraic Geometry
pdf ↗ page arXiv:2611.07202 ↗

We consider the category of sheaves on a manifold with microsupport in a singular conic Lagrangian, and more generally microsheaves on the Lagrangian skeleton of a Weinstein manifold with stop, which are equivalent to ind-completions of partially wrapped Fukaya categories. Compact objects in these categories are obtained as wrappings of constructible sheaves with perfect stalks; they correspond to objects built from exact Lagrangians in finitely many steps. We give a microlocal criterion on constructible (micro)sheaves with perfect (micro)stalks which guarantees that the corresponding wrapped sheaves induce integral kernels that preserve compact objects of these categories. This generalizes a result of Gao on Lagrangian correspondences between fully wrapped Fukaya categories.

comments
32 pages
fields
Symplectic Geometry, Algebraic Geometry
pdf ↗ page arXiv:2611.07466 ↗

Let $(M,ω_M)$ be closed symplectic manifold containing a closed Lagrangian submanifold $L$ with $ω_M$ vanishing on $π_2(M,L)$. We prove that there is a nonconstant $J$-holomorphic plane $u:\mathbb{C}\rightarrow M\times \mathbb{T}^2$ for any almost complex structure $ J$ on $M\times \mathbb{T}^2$ tamed by $ω_M\oplusω_{\text{std}}$.

comments
4 pages. Comments are welcome!
fields
Symplectic Geometry
pdf ↗ page arXiv:2611.07799 ↗

This paper is motivated by the recent discovery, in symplectic topology, of a relation (partly proved, partly conjectural) between quantum connection and quantum Steenrod operations. We introduce an abstract TQFT framework, based on certain features of the definition of deformed symplectic cohomology, in which such a relation holds generally.

fields
Symplectic Geometry
pdf ↗ page arXiv:2611.07943 ↗

We describe the shifted Lagrangian geometry of curves in Radko surfaces, through their global integrations and their homotopy Poisson algebras. We construct the adjoint symplectic groupoid of the surface explicitly, show that admissible curves integrate to embedded Lagrangian subgroupoids, which are 1-shifted Lagrangians, and characterize when their shifted intersections are clean. For a component of the degeneracy locus, a canonical Hamiltonian flow on its shifted self-intersection recovers the modular period. On the infinitesimal side, the homotopy Poisson algebra of a connected admissible curve not contained in the degeneracy locus is $L_\infty$-isomorphic to its underlying cochain complex, and its $L_\infty$-isomorphism class is determined by the total contact multiplicity. The brackets of arity at least two can be removed by a change of tubular neighbourhood exactly when all contacts are transverse. At tangential contacts, the binary bracket retains the modular vector field. We establish this distinction using a general characterization of the $L_\infty$-isomorphisms induced by changes of tubular neighbourhood, valid for arbitrary coisotropic submanifolds.

fields
Symplectic Geometry, Differential Geometry, Quantum Algebra
pdf ↗ page arXiv:2611.08248 ↗

Geometric Topology4

We give handlebody decompositions of the complements of the cyclic multisections in $Σ_g\times S^2$ constructed in~\cite{AB}. The 2-handle framings are induced by the product coordinates. We compute the integral homology of the complements from their handle chain complexes, including the torsion subgroup of $H_2$. For the degree pairs $(3,3)$ and $(2,4)$, we specify surface diffeomorphisms inducing the maps on first homology prescribed in~\cite{AY} and recover the integral first homology groups $(\Z/3)^2\oplus\Z/(3m)$ and $(\Z/2)^2\oplus\Z/(4m)$, respectively, for $m\geq1$. For the symmetric $m=1$ sum, we also give a simultaneous surface diagram of the relative product attachments and identify their product framings with the surface framings. The resulting symplectic sums are aspherical and symplectically minimal, and have the rational cohomology ring of $S^2\times S^2$.

comments
39 pages, 27 figures
fields
Geometric Topology, Algebraic Topology
pdf ↗ page arXiv:2611.06949 ↗

Given a stratum $\mathcal{H}(μ)$ of meromorphic one-forms on a Riemann surface with prescribed orders of zeros and poles, the residual map assigns to each differential the collection of its residues at the poles. In this paper, we prove that the residual map is a locally trivial topological fibration over each open flat of the stratification induced by the resonance arrangement. To give an algebro-geometric proof of this result, we introduce a refined compactification of the projectivized stratum $\mathbb{P}\mathcal{H}(μ)$, namely the moduli space of residue-aligned multi-scale differentials $\operatorname{RAMS}(μ)$. We prove that the projectivized residual map extends to $\operatorname{RAMS}(μ)$ as a stratified fibration over the projectivized resonance arrangement, thereby yielding a resolution of the projectivized residual map. This approach naturally extends to the analogue of the residual map for strata of $k$-differentials. Complementing the algebro-geometric approach, we also develop a flat-geometric point of view. In genus zero, each isoresidual fiber carries a canonical higher-dimensional analogue of a translation structure. These fibers admit a metric completion whose boundary has a modular interpretation in terms of the WYSIWYG compactification, for which we provide explicit local models. Isoresidual fibers admit a polyhedral cell decomposition whose geometry varies linearly with the underlying configuration of residues in period coordinates. Degenerations of these cells occur precisely when resonance conditions are satisfied, providing a geometric proof that the residual map is a stratified fibration. Finally, certain isoresidual fibers admit natural interpretations as Hurwitz spaces of branched covers, yielding unexpected homeomorphisms between moduli spaces of covers with prescribed ramification profiles.

comments
54 pages
fields
Geometric Topology, Algebraic Geometry
pdf ↗ page arXiv:2611.07499 ↗

Using dehydration signatures as an encoding, we train GPT-2-style transformers to generate triangulations of 3-dimensional manifolds. Our main runs each use approximately 3 exaFLOPs, spanning a 200 million parameter model trained on 60 million sphere triangulations to an 11 million parameter model trained on 1.1 billion sphere triangulations. The optimal model has 10 blocks, width 640 and generates dehydration signatures of closed manifolds of various topological types at a rate of 6.9%, 82% of which are spheres. To investigate how our method scales, we use the technique of IsoFLOP curves on 15-tetrahedron triangulations of the 3-sphere, over a compute range of 6 to 3000 petaFLOPs, and observe neural scaling behaviour. We extract a parameter scaling exponent of $α= 0.49$ ($[0.48, 0.50]$ stat, $\pm 0.02$ syst), and an optimal ratio of 234 tokens per parameter for training. In addition, we present a number of smaller experiments, including evidence that transformers are able to learn to count degree-one edges near-perfectly, but get progressively worse when counting edges with higher degrees.

comments
21 pages + 3 pages appendix, 7 figures, 10 tables
fields
Geometric Topology
pdf ↗ page arXiv:2611.07514 ↗

We study the counting function $\mathcal{P}^m(X,x)$ for $m$-multigeodesics of length at most $x$ on a compact hyperbolic surface $X$ where $m$ is a positive integer. Building on the Selberg-Delange method, originally developed for counting integers with exactly $m$ prime factors in the interval $[1,x]$, we prove a large $x$ asymptotic for $\mathcal{P}^m(X,x)$. We show that for fixed $m$ as $x\to\infty$ the asymptotic is independent of the surface, whereas for $\frac{m}{\log x}$ bounded away from zero, the asymptotic depends on an infinite Euler product over the lengths of primitive closed geodesics. We also study the distribution of the function $ω_1$, defined on multigeodesics $Γ$ as $ω_1(Γ)=k-1$ for $Γ=(γ_1,...,γ_k)$ a $k$-multigeodesic. We show that as $x\to\infty$, the distribution of $ω_1$ in the set of multigeodesics of length $\leq x$ is close to that of a Poisson random variable with parameter $\log x$. Both of these theorems rely on studying the generating function of the random variable $ω_1$ which we relate to the Selberg zeta function.

comments
23 pages, 1 figure
fields
Geometric Topology, Number Theory, Spectral Theory
pdf ↗ page arXiv:2611.08165 ↗

Operator Algebras4

We study weak almost periodicity and left uniform continuity for Hopf-von Neumann algebras and locally compact quantum groups. For a Hopf-von Neumann algebra $(M,Γ)$, we introduce the space ${\rm WAP}_{{\rm Bessel},π_u}(M_*)$ through norm convergence conditions on rows and columns of slices associated with Bessel families in the universal representation. We prove that this space is the largest $C^*$-subalgebra of ${\rm WAP}(M_*)$ when separately continuous lift of the co-multiplication is multiplicative, and that the two spaces coincide when $M$ is commutative. Furthermore, the algebra $\{x\in C_b(\mathbb{G}): Γ(x)\in M(C_0(\mathbb{G})\otimes C_b(\mathbb{G}))\}$ is characterized as the greatest left introverted $C^*$-subalgebra of ${\rm LUC}(L^1(\mathbb{G}))$ satisfying a left Bessel condition on $L^2(\mathbb{G})$. We also show that ${\rm LUC}(L^1(\mathbb{G}))$ is a $C^*$-algebra whenever the center of $C_0(\mathbb{G})$ contains an element on which the bounded counit is non-zero.

fields
Operator Algebras, Functional Analysis
pdf ↗ page arXiv:2611.07168 ↗

For an action of the integers on a C*-algebra, we show that the only possible values for the Rokhlin dimension (with commuting towers) are $0,1,2$ and $+\infty$. We study these values for certain classes of C*-algebras, and show that there is an automorphism of the Jiang–Su algebra with the property that its Rokhlin dimension with and without commuting towers are distinct.

comments
Version 1. Comments Welcome
fields
Operator Algebras
pdf ↗ page arXiv:2611.07556 ↗

This is a nearly complete draft of our paper in which we attempt to do three things. First, we show that when formulated in the context of reproducing kernel correspondences over Rieffel's category, Normod-$M$, the Kolmogorov decomposition/representation theorem of Ball, Marx, and Vinnikov is functorial in $M$. Second, we investigate relations between our work and the work of Clouâtre and Hartz concerning multipliers of reproducing kernels. This investigation leads to interesting questions about the geometry of the matricial unit disc associated with a correspondence over $M$. And third, we describe how some of the recent work of Salomon, Shalit and Shamovich may be formulated in our setting. This leads to the notion of subvarieties, their quotients, and their relation with the structure of reproducing correspondences.

comments
54 pages
fields
Operator Algebras, Functional Analysis
pdf ↗ page arXiv:2611.07700 ↗

Let $M$ be a semifinite von Neumann algebra with a faithful normal semifinite trace $τ$, and let $Φ$ be an Orlicz function. We prove that the noncommutative Orlicz space $L^Φ(M,τ)$ is contained in $M$ if and only if $M$ is purely atomic and there is $δ>0$ such that $τ(e)\geqδ$ for every nonzero projection $e\in M$. We also prove that every left $M$-module map on $L^Φ(M,τ)$ is bounded. As a consequence, a left ideal of $M$ contained in $L^Φ(M,τ)$ and dense in it admits an $M$-invariant algebraic complement if and only if it is equal to $L^Φ(M,τ)$. These results extend a theorem of Fidaleo and Zsidó from $L^2$ to all noncommutative Orlicz spaces.

comments
12pp. Comments are welcome
fields
Operator Algebras
pdf ↗ page arXiv:2611.07978 ↗

Rings and Algebras4

The article studies the notion of the structure graph of an associative ring with unity. It is known that the structural graph of a ring with $n$ elements has at least $n$ connected components. The article investigates when this lower bound is attained. A hypothesis is stated that the only rings of order $n$ having exactly $n$ connected components are the rings $\mathbb{Z}_2$, $\mathbb{Z}_3$, $\mathbb{Z}_7$, $GF(4)$, and their direct sums. The hypothesis is proved for all rings except residue rings modulo a prime of the form $7+12k$, $k\geq181$, and Galois fields of characteristic~2 with extension degree $\geq17$.

comments
11 pages, 2 figures
fields
Rings and Algebras
pdf ↗ page arXiv:2611.07521 ↗

We study local and $2$-local derivations of finite-dimensional Zinbiel algebras. Over any field of characteristic zero, every Zinbiel algebra $A$ with $A^2 \neq 0$ and $\dim A \leq 4$ admits a proper local derivation. The proof uses a filtration argument. We exhibit two five-dimensional Zinbiel algebras with nonzero multiplication whose derivations are all central. Every local derivation on either algebra is therefore a derivation, so the minimum dimension for local rigidity with nonzero multiplication is five. Annihilator-valued interpolation gives a general construction of proper local derivations. When $A^3 = 0$, the multiplication tensor describes all derivations and local derivations. For a finite-dimensional nilpotent algebra $A$ over any field, we show that every $2$-local derivation is a derivation if and only if $\dim(A/A^2) \leq 1$. The proof combines the annihilator interpolation of Sh. Ayupov, K. Kudaybergenov and B. Omirov with a one-generator argument. When all derivations are central, every $2$-local derivation is obtained by composing a homogeneous map $A/A^2 \to \text{Ann}(A)$ with the quotient map, and every such composition is a $2$-local derivation. The five-dimensional examples extend to locally rigid Zinbiel algebras with nonlinear $2$-local derivations in every dimension at least five. We also show that derivations given by left multiplication form a Lie ideal of nilpotency class at most two and give sufficient conditions for the existence of left outer derivations. This ideal is zero in our locally rigid examples.

fields
Rings and Algebras
pdf ↗ page arXiv:2611.07863 ↗

For any $m$ pairwise commuting matrices $\{A_i\}_{i=1}^{m}\subset \M_n(\F)$, we prove $\dim_\F \F[A_1,\cdots, A_m]\leq (2^m-1)(n-1)+1$, answering Bergman's question on the existence of a linear bound for a fixed number $m$.

fields
Rings and Algebras
pdf ↗ page arXiv:2611.08135 ↗

Let E be an arbitrary graph, K a field, and D the set of boundary paths of E consisting of both infinite paths and finite paths ending at singular vertices. The space KD of K-valued functions on D is a right module over the Leavitt path algebra LK(E). For a non- empty set S ? D that is strongly connected and closed under adding and deleting initial segments, we study the submodule N_S = 1_S.LK(E) generated by the indicator function of S. Let S_0 be the set of vertices lying on the paths in S. We show that N_S is simple if and only if S contains no breaking vertex of the hereditary saturated set H = E0 nS0. In particular, when S contains only infinite paths or when E is a row- finite graph, NS is a simple module. Up to isomorphism, the simple module NS depends only on S_0. When S is an infinite set of paths, the simple module NS is neither graded nor projective. The left module LK(E)1_S, obtained by twisting the action through the involution of LK(E), is simple exactly when NS is, and, when S is infinite, it is not isomorphic to any of the Chen simple modules. For the rose graph R_n with n ? 2 petals, we show that LK(R_n)1_S is isomorphic to one of the twisted Chen modules constructed by Nam, Srivastava and Vien, and to none of the other simple modules in their list. Moreover, when S is not a single vertex, the simple module NS is finitely presented if and only if only finitely many edges of E have both source and range in S_0; in particular, every simple module N_S is finitely presented when E is a finite graph. Finally, we determine when the analogous module generated by an indicator function over the Cohn path algebra CK(E) is simple. 1.

comments
29 pages
fields
Rings and Algebras
pdf ↗ page arXiv:2611.08344 ↗

Commutative Algebra3

Let $R$ be a commutative ring with exactly $n\ge 2$ maximal ideals and let $k\ge 2$. We study the minimal $k$-uniform comaximal hypergraph $\mathcal{H}_k(R)$, whose vertices are the nonunits outside the Jacobson radical and whose edges are the irredundant $k$-element comaximal families. For $2\le k\le n$ we show that $\mathcal{H}_k(R)$ is a blow-up of the hypergraph of irredundant $k$-coverings of $\{1,\dots,n\}$, and that its reduced form reconstructs the maximal ideals up to permutation; for $k>n$ there are no edges. We obtain a general description of $\mathrm{Aut}(\mathcal{H}_k(R))$ and characterize isomorphisms of such hypergraphs. For $3\le k\le n$ its automorphism group properly contains that of the comaximal graph, and for finite rings with $n\ge 3$ and every $2\le k\le n$, $\mathcal{H}_k(R)$ has the same isomorphism invariants as the comaximal graph, namely $|J(R)|$ and $R/J(R)$. The case $k=2$ recovers and extends the automorphism results for comaximal graphs obtained previously. We also compare the minimal and non-minimal comaximal hypergraphs and identify a condition under which the latter has the same automorphism group and isomorphism invariants as the comaximal graph. The same method determines the vertex set and the automorphism group of the $k$-zero-divisor hypergraph of a product of integral domains, and shows that for finite products of finite fields this hypergraph determines the ring.

comments
13 pages
fields
Commutative Algebra, Combinatorics
pdf ↗ page arXiv:2611.07992 ↗

The comaximal graph $\mathcal{G}(R)$ of a commutative ring $R$ has as vertices the nonunits outside the Jacobson radical, two being adjacent if they generate $R$ as an ideal. For rings with $k\ge 4$ maximal ideals, only the bounds $2\leγ(\mathcal{G}(R))\le k$ were known, to our knowledge. We prove $γ=γ_t=γ_c=k$ whenever $3\le k<\infty$, without finiteness or Artinian hypotheses, classify the minimum dominating sets, and determine the domatic number. Independent dominating sets correspond to maximal intersecting families of subsets of $\{1,\dots,k\}$; we obtain a closed formula for the independent domination number of every finite non-local ring, which can be exponentially larger than the domination number. The proofs rest on the class sizes of the blow-up description of $\mathcal{G}(R)$ due to Moconja and Petrović, which we determine for every ring with finitely many maximal ideals, finite or infinite; with them, domination problems on $\mathcal{G}(R)$ become weighted extremal problems on subsets of $\{1,\dots,k\}$, and the formula for the independent domination number follows from the multiplicative form of these sizes. We also treat rings with infinitely many maximal ideals, graph polynomials, and lattices whose maximal ideals are prime.

comments
12 pages
fields
Commutative Algebra, Combinatorics
pdf ↗ page arXiv:2611.08022 ↗

Let $D$ be an integral domain and let $S\subseteq D\setminus\{0\}$ be multiplicative. We study the condition that, for each nonzero $a,b\in D$, some $(a^n,b^n)_v$ is $S$-principal. A power estimate extends this binary condition to arbitrary finite families and permits synchronization of exponents. We describe the obstruction to descent from the AGCD domain $D_S$ by bounded $S$-torsion and by a finiteness condition on intersections of principal ideals. For an almost Prüfer $v$-multiplication domain, the almost $S$-GCD property, the AGCD property of $D_S$, and torsion of the $S$-class group are equivalent. In Krull domains this gives a criterion in terms of the divisor classes of height-one primes meeting $S$, and finite rational class rank reduces the multiplicative set to one element. We prove conductor transfer without a principal-contraction hypothesis, polynomial transfer in several settings, a normalization homeomorphism away from $S$, and a finite-principal-cover criterion with a Picard-group obstruction. Explicit examples distinguish localization, bounded descent, and the global and sublocal conditions.

comments
20 pages
fields
Commutative Algebra
pdf ↗ page arXiv:2611.08340 ↗

General Mathematics3

We study the radius of absolute monotonicity of rational approximations to the exponential of the form \[ φ(z)=\frac{P(z)}{(1-z/b)^s}, \qquad b>0,\qquad °P\le s, \] with approximation orders two and three. For second-order approximation, we prove that the maximal radius in degree $s=5$ is $10$ and determine the unique extremizer. Hence the conjectured value $2s$ by Lóczi and Ketcheson (LMS Journal of Computation and Mathematics, 17 (2014), no.~1, 159–205) remains valid in this case. In contrast, for every integer $s\ge6$ we construct explicitly a rational function with rational coefficients and exact radius \[ 2s+\frac{4(s-3)}{s^2-3s+4}>2s. \] For third-order approximation, we construct, for every $s\ge5$, an explicit rational family with exact radius \[ 2s-1>s-1+\sqrt{s^2-1}. \] These constructions disprove the conjectures of Lóczi and Ketcheson concerning rational functions with a single repeated pole.

comments
23 pages
fields
General Mathematics
pdf ↗ page arXiv:2611.06941 ↗

Mahjong is a traditional tile-based game popular in East and Southeast Asia. Combinatorial study of Mahjong hands has been a topic of interest in recreational mathematics, with the Nine Gates hand being a well-known example. This single-suit Mahjong hand $\{1,1,1,2,3,4,5,6,7,8,9,9,9\}$ consists of $13$ tiles of a single suit ranked $1$ through $9$, arranged such that drawing any tile from that suit completes a winning hand of a pair and $4$ melds, regardless of the tile's rank. Folklore has circulated that Nine Gates is the unique single-suit hand with this $9$-sided-wait property. While prior works have verified this uniqueness by exhaustive computer enumeration across different finite configurations, we present a general classification of full-sided waits for Mahjong variants with $3n$ ranks and $n+1$ melds. For every $n\geq2$, we report exactly $8$ full-sided waits, of which Nine Gates is the unique live one. Notably, our proof requires no exhaustive enumeration or any prior knowledge of the classification result. To address the problem in its right generality, a comprehensive combinatorial framework is developed, allowing the number of ranks, the number of melds, and the upper bound on tile multiplicities to vary independently. Reliability of our results is partially supported by a manually curated ongoing formalization in Lean 4.

code
github.com/sun123zxy/ninegates-lean4 ∎ Lean project (lakefile.toml, lean-toolchain) builds, no sorry at e45fd51, checked 2026-11-17 build details
comments
19 pages, 5 tables. Lean 4 formalization: https://github.com/sun123zxy/ninegates-lean4 . Interactive visualization: https://sun123zxy.github.io/ninegates-viz/
fields
General Mathematics
pdf ↗ page arXiv:2611.06944 ↗

We outline the algorithmic advances behind our recent computation of the number 8364925094730088313 of ways to combine eleven 2 x 4 LEGO bricks.

fields
General Mathematics
pdf ↗ page arXiv:2611.06986 ↗

Quantum Algebra3

We obtain classification results for rational vertex operator algebras generated in low conformal weights. More precisely, we prove that every rational vertex operator algebra of OZ type strongly generated by its conformal weight two subspace $V_2$ is isomorphic to a finite tensor product of Virasoro minimal models. We also prove that every rational vertex operator algebra of CFT type generated as a vertex algebra by its conformal weight one subspace $V_1$ is isomorphic to a finite tensor product of simple affine vertex operator algebras associated with finite-dimensional simple Lie algebras at positive integral levels. Consequently, Zhu's conjecture that rationality implies $C_2$-cofiniteness holds for both classes, and every such rational vertex operator algebra is regular.

fields
Quantum Algebra
pdf ↗ page arXiv:2611.06983 ↗

In this paper, we give diagrammatic descriptions of the quantum subgroups coming from conformal embeddings of the form $\mathcal{V}(\mathfrak{g}), \check{h}) \subset \mathcal{V}(\mathfrak{so}_{\dim \mathfrak{g}}, 1)$ for $\mathfrak{g}\in \left\{\mathfrak{so}_N, \mathfrak{sp}_{2N}\right\}$, as well as the conformal embeddings of the form $\mathcal{V}(\mathfrak{sp}_{2r + 2}, 2r) \subset \mathcal{C}(\mathfrak{so}_{\dim \mathfrak{sp}_{2r + 2}}, 1)$, extending the techniques of Edie-Michell and Snyder which applied to conformal embeddings of the form $\mathcal{V}(\mathfrak{sl}_r, r) \subset \mathcal{V}(\mathfrak{so}_{\dim \mathfrak{sl}_r}, 1)$. Our main conceptual innovation is the use of fermion condensation to construct a continuous family of planar superalgebras, which we call the Clifford BMW supercategories. The tower of endomorphism algebras, $CW_n(σ, ε)$ is a combination of Birman Wenzl Murakami algebras and Clifford algebras, much in the spirit of the Hecke Clifford algebras. This tower of algebras appears to be new, and already yields some exciting combinatorics.

fields
Quantum Algebra, Representation Theory
pdf ↗ page arXiv:2611.07221 ↗

We study the twisted Yangian associated with the quasi-split symmetric pair $(\mathfrak{o}_{2n}, \mathfrak{o}_{n-1}\oplus \mathfrak{o}_{n+1})$ of type DI. Working in the transposed $R$-matrix presentation of arXiv:2607.14692, we show that the generating matrix $S(u)$ admits no Gaussian decomposition. We decompose instead the matrix $\widehat{S}(u) = G\,S(u)$, where $G$ is the constant term of $S(u)$. A chain of quasi-determinantal embeddings lowers the rank within this family of pairs and ends at $(\mathfrak{o}_4,\mathfrak{o}_3)$. Together with a direct study of a degenerate block of type AI, this gives a presentation of the special twisted Yangian $SY^{\mathrm{tw}}(\mathfrak{o}_{2n},\mathfrak{o}_{n-1}\oplus \mathfrak{o}_{n+1})$ by Drinfeld type current relations. The two fork nodes of the Dynkin diagram carry Drinfeld's presentation of $Y(\mathfrak{sl}_2)$. All other relations are those of split type DI, except that the Serre relation with a doubled fork node has zero right hand side. For $n\ge3$ we obtain an explicit isomorphism between the $R$-matrix presentation of Guay and Regelskis and the Drinfeld presentation of Lu and Zhang. The Drinfeld generators are expressed through the Gaussian series of $\widehat{S}(u)$. With quasi-split type AIII (Lu and Zhang), this completes the explicit comparison for the classical quasi-split types with a nontrivial diagram involution.

comments
25 pages. Companion to arXiv:2607.14692
fields
Quantum Algebra, Representation Theory
pdf ↗ page arXiv:2611.08357 ↗

General Topology2

We study Hamel character topologies on the real line in two parts. Part I treats the full real-orbit (TVG) and the rational-parameter (TVGQ) dual pairs. A divisibility-chain criterion decides compatibility on the rationals and on Q/Z. Compatible subsets of one chain have a compatible join, but two different chains need not: an explicit Chinese remainder construction gives two compatible topologies with an incompatible join. This answers the compact-union question negatively, disproves the Greatest Compatible TVG Conjecture for every nonzero rational-valued Hamel functional, and shows that the corresponding rational and real dual pairs admit no Mackey topology. We classify additive homomorphisms and isomorphisms of the real-valued observation spaces and their automorphism groups, prove that both prime square-denominator topologies are non-nuclear, classify sequences converging for every coefficient functional, and show that the real parameters of continuous characters of the integers with the mixed-prime metric form a set of Hausdorff dimension zero. Part II studies real vector spaces whose Hausdorff additive group topology makes every fixed rational scalar multiplication continuous (TVGQ). A rational-valued Hamel functional and a rational parameter space S generate precompact, modular and real-valued observation topologies whose exact scalar ring is the multiplier ring of S. We compute their completions and character groups, classify the precompact TVGQ topologies, characterize universal separating sequences, show that local compactness and sigma-compactness force the scalar ring to be a field, classify isomorphisms within each modular family (exactly 2^c types), identify greatest-compatible existence with the unrestricted Mackey problem, and prove that all compatible topologies for the rational Hamel pair are non-Baire.

fields
General Topology
pdf ↗ page arXiv:2611.07052 ↗

An undirected connected graph can be interpreted as a metric space by considering the path distance between its vertices. Extending this perspective, we explore how walks in the graph, viewed as sequences of subsets of vertices $V$, fit into a more general framework where the path distance emerges as a special case of broader metrics that define hypertopologies on the space of subsets of $V$. In this paper, we propose a general framework to address this issue, motivated by the extension of recent reinforcement learning algorithms on graphs through Lipschitz function extensions. Our approach demonstrates how models for dynamical systems and related machine learning techniques can be adapted to (quasi-pseudo-metric) hyperspaces, offering a novel perspective for the analysis and modeling of dynamical systems. Finally, we illustrate the practical relevance of our theoretical contributions through a concrete graph optimization example, highlighting the utility and effectiveness of our proposed approach.

comments
22 pages, 3 figures
fields
General Topology
pdf ↗ page arXiv:2611.07860 ↗

History and Overview2

This article presents web-based applets designed primarily to educate future mathematics teachers in the field of numerical algorithms and, secondarily, to serve high school students themselves. The applets allow users to explore the behaviour of basic numerical algorithms as applied to fundamental mathematical topics taught at high schools. The advantage is that no software is required, and students can intuitively test how a given method works by adjusting specific problem parameters, not only by viewing a list of approximate solutions, but especially through graphical output. Furthermore, they have access to multiple different methods for comparison.

comments
12 pages, 7 figures
fields
History and Overview, Numerical Analysis
pdf ↗ page arXiv:2611.06938 ↗

Cographs, the $P_4$-free graphs, are exactly the graphs obtained from a single vertex by repeated disjoint union and complementation. They contain the threshold graphs, $\mathcal{C}$-graphs and form one of the primary areas for spectral graph theory. This survey is organized by matrix: after a preliminary section, there is one section for each of the adjacency, Laplacian, signless Laplacian, distance, eccentricity, Seidel, distance Laplacian and distance signless Laplacian matrices. Within each section we collect what is known for general cographs, for the subclass of $\mathcal{C}$-graphs, and where relevant, for threshold graphs. We highlight the results from the recent years, including the foundational work of Mohammadian et al., Jacobs et al., Ghorbani, Li et al., Niu et al., Lou et al., Wang at al., Mandal et al. Two structural facts are used throughout: every connected cograph has diameter at most $2$, which ties all four distance-type matrices to the adjacency and Laplacian matrices, and the partition of the vertex set into twin classes is equitable for every one of the eight matrices, giving a uniform description of each spectrum (for the Laplacian this is the twin-reduction theory of Barik–Reddy, which we present in detail). Besides surveying the literature, we record several consequences, and mark them as such: an exact decomposition of the eccentricity matrix of a connected cograph, yielding an eigenvalue-free interval for the eccentricity spectrum of all connected cographs; the full distance Laplacian spectrum and its extremal eigenvalues; an eigenvalue-free interval for the distance signless Laplacian valid for every connected graph of diameter at most $2$; an interlacing theorem tying the distance spectrum to the adjacency spectrum. A comparison table across all eight matrices and a list of open problems close the paper.

fields
History and Overview, Combinatorics
pdf ↗ page arXiv:2611.06981 ↗

K-Theory and Homology1

We construct higher additive noncommutative motives of stable $(\infty,n)$-categories, extending the additive motives of Blumberg, Gepner, and Tabuada and the secondary theory of Mazel-Gee and Stern. We formulate a notion of exact sequences and establish the universal property of the resulting additive motives. We give two equivalent constructions: a localization of enriched presheaves and a weighted Waldhausen $S$-construction on complete exact flags. We then compute their Hom objects, obtaining ordinary K-theory at the first level and lower additive motives recursively at higher levels. Our results recover the classical and secondary theories, provide coherent multiplicative structures, and yield an endofunctor on the category of Stefanich rings. In short, higher motives are higher K-theory.

fields
K-Theory and Homology, Algebraic Topology, Category Theory
pdf ↗ page arXiv:2611.07262 ↗

Spectral Theory1

Pure point spectral convergence is generally unavailable for sparse graph sequences, where Laplacian eigenvalues may remain unstable under graph growth. This poses challenges for analyzing spectral operators underlying many graph-based learning methods, including graph neural networks. We identify a class of sparse bipartite graph sequences for which eigenvalue-level spectral convergence can be recovered. Focusing on graphs with strong degree heterogeneity, we exploit the convergence properties of their line-graph sequences. Using a spectral correspondence between bipartite graph Laplacians and line-graph adjacency operators, we transfer convergence guarantees from dense line-graphs back to the original sparse graphs. This framework yields eigenvalue-level spectral convergence for a class of deterministic sparse graph sequences. We further analyze superlinear preferential attachment trees, establishing convergence to an infinite-dimensional limit operator. Numerical experiments empirically support our spectral convergence results.

fields
Spectral Theory
pdf ↗ page arXiv:2611.08246 ↗

Revised papers244

For $\mathbb{K}=\mathbb{R}$ or $\mathbb{C}$ and every $n\ge 1$ we construct two $n\times n$ matrices $A$ and $B$ such that the orbit of every vector $p\in\mathbb{K}^n$ with $p_1\ne 0$ under the semigroup generated by $A$ and $B$ is dense in $\mathbb{K}^n$; the remaining vectors lie in an invariant hyperplane, so the set of vectors with dense orbit is exactly the complement of a hyperplane. Thus two generators always suffice for a hypercyclic semigroup of matrices, in contrast with the commutative case, where $n+1$ generators are needed on $\mathbb{C}^n$. We give a transparent reformulation of the hypotheses, a criterion for the density of $\{a^mb^n\}$ that makes them checkable, and explicit examples with complete proofs. As applications we obtain affine pairs on $\mathbb{K}^n$ all of whose orbits are dense, and, for every $n$, a pair of real matrices with an orbit that is somewhere dense but not dense in $\mathbb{R}^n$.

comments
This is a revised version of a published article
journal
M. Javaheri, Semigroups of matrices with dense orbits, Dyn. Syst. 26 (2011), no. 3, 235--243
fields
Dynamical Systems, Functional Analysis
pdf ↗ page arXiv:0905.1311 ↗

Let $θ(x)=(x-1)/3$ if $x\ge1$ and $θ(x)=2x/(1-x)$ if $0\le x<1$. We conjecture that the $θ$-orbit of every nonnegative rational number ends in $0$. A weaker conjecture asserts that no map in the semigroup generated by $3x+1$ and $x/(x+2)$, other than the generators' powers, has a positive rational fixed point; equivalently, the only matrices in the semigroup generated by $\left(\begin{smallmatrix}3&1\\0&1\end{smallmatrix}\right)$ and $\left(\begin{smallmatrix}1&0\\1&2\end{smallmatrix}\right)$ with an integer eigenvalue are their pure powers. We prove sharp two-sided bounds for the smallest eigenvalue of every word in the more general semigroups attached to the maps $\varphi_{a,b}$, which explain why the pair $\{2,3\}$ is distinguished: it is the only pair with $(a-1)(b-1)=2$. Combining these bounds with congruences for the trace, we determine the possible integer eigenvalues of words, and we prove that $θ$ has no rational $1$-circuits, no rational $2$-circuits and no even rational circuits. We show that, for each $k$, a random $k$-syllable word with exponents at most $M$ has an integer eigenvalue with probability at most $2k(k+\log_2 2k)/M$. Finally, we verify by computer that $θ$ has no positive rational cycle of length at most $36$.

comments
This is a revised and extended version of a published article
journal
M. Javaheri, A Collatz-type conjecture on the set of rational numbers, J. Number Theory 132 (2012), no. 11, 2582--2594
fields
Number Theory
pdf ↗ page arXiv:1010.3692 ↗

We show that the semigroup of real linear fractional transformations of a proper subinterval of the real line does not contain a dense subsemigroup generated by two elements, while it does contain dense subsemigroups generated by three elements. For the semigroup generated by the two maps $x\mapsto 1+a/x$ and $x\mapsto x/b$ (the only pair not excluded by elementary arguments) we compute the closure exactly, and we describe all of its orbit closures in the plane. Consequently the semigroup of $2\times2$ matrices with nonnegative entries has no dense $2$-generator subsemigroup but has dense $3$-generator ones. In contrast, we prove a general density criterion for subsemigroups of $M_2(\mathbb{R})$ and $M_2(\mathbb{C})$ generated by three explicit matrices, and use a word identity of length seven to produce, for both $\mathbb{R}$ and $\mathbb{C}$, uncountable families of pairs of matrices that generate dense subsemigroups; a fully explicit rational example is $\left(\begin{smallmatrix}1&1/2\\1&0\end{smallmatrix}\right)$ and $\left(\begin{smallmatrix}1&0\\0&-8/3\end{smallmatrix}\right)$. Thus the minimal number of generators of a dense subsemigroup is $3$ for the nonnegative cone and $2$ for $M_2(\mathbb{R})$ and $M_2(\mathbb{C})$.

comments
This is a revised and extended version of a published article
journal
M. Javaheri, Dense 2-generator subsemigroups of 2x2 matrices, J. Math. Anal. Appl. 387 (2012), 103--113
fields
Number Theory, Dynamical Systems
pdf ↗ page arXiv:1104.5054 ↗

Let $\mathbb{K}=\mathbb{R}$ or $\mathbb{C}$. We prove that for every $n\ge 1$ there is a pair of $n\times n$ matrices with entries in $\mathbb{K}$ that generates a dense subsemigroup of the space $M_n(\mathbb{K})$ of all $n\times n$ matrices. Since the powers of a single matrix are never dense, two generators is optimal. Equivalently, there is a $2$-generator semigroup of linear maps whose action on $(\mathbb{K}^n)^n$ is topologically transitive, whereas no semigroup of linear maps acts topologically transitively on $(\mathbb{K}^n)^{n+1}$. In fact, for such a semigroup every orbit closure in $(\mathbb{K}^n)^k$, for every $k\ge1$, is as large as linear algebra allows. Along the way we give a short self-contained proof of the density of explicit $2$-generator subsemigroups of $GL_2^+(\mathbb{R})$ and of $M_2(\mathbb{R})$, based on a simple criterion for closed subsemigroups of $SL_2(\mathbb{R})$.

comments
This is a revised and extended version of a published article
journal
M. Javaheri, Maximally transitive semigroups of nxn matrices, J. Math. Anal. Appl. 401 (2013), 743--753
fields
Dynamical Systems, Functional Analysis
pdf ↗ page arXiv:1201.0192 ↗

(Due to the limit on the number of characters for an abstract set by arXiv, the full abstract can not be displayed here. See the abstract in the paper.) We study dispersive equations with a time non-homogeneous modulation acting on the linear dispersion term. As primary models, we consider the Korteweg-de Vries equation (KdV) and related equations such as the Benjamin-Ono equation (BO) and the intermediate long wave equation (ILW), imposing certain irregularity conditions on the time non-homogeneous modulation. In this work, we establish phenomena called regularization by noise in three-folds: (i) When the modulation is sufficiently irregular, we show that the modulated KdV on both the circle and the real line is locally well-posed in the regime where the (unmodulated) KdV equation is known to be ill-posed. In particular, given any $s \in \mathbb R$, we show that the modulated KdV on the circle with a sufficiently irregular modulation is locally well-posed in $H^s(\mathbb T)$. Moreover, by adapting the $I$-method to the current modulated setting, we prove global well-posedness of the modulated KdV in negative Sobolev spaces. (ii) It is known that certain (semilinear) dispersive equations such as BO and ILW exhibit quasilinear nature. We show that sufficiently irregular modulations make the modulated versions of these equations semilinear by establishing their local well-posedness by a contraction argument, providing local Lipschitz continuity of the solution map. (iii) We also prove nonlinear smoothing for these modulated equations, where we show that a gain of regularity of the nonlinear part becomes (arbitrarily) larger for more irregular modulations. As applications of our approach, we also include further examples.

comments
94 pages. Updated funding info. To appear in Mem. Eur. Math. Soc
fields
Analysis of PDEs, Probability
pdf ↗ page arXiv:1406.7675 ↗

We study the Hausdorff measure of noncompactness associated with compatible metrics on Fréchet spaces, with the aim of extending its Banach-space properties. After recalling the characterization by complete translation-invariant metrics with convex balls, we calculate the scalar Hausdorff measure for a standard weighted fractional metric. For bounded nonprecompact sets its value is the unique root of an equation involving a finite seminorm gauge determined by the covering radius. This yields an actual Hausdorff-condensing criterion for compact perturbations of maps having separate contraction constants at the seminorm levels. On countable products of Banach spaces we obtain an exact coordinate formula and a sharp scalar contraction bound on bounded domains, without requiring a uniform gap between the coordinate constants and one for strict condensation. For closed subspaces, an exact finite-projection formula retains the compatibility constraints on covering centers. Explicit examples show why separately computed seminorm measures cannot generally be substituted into the metric formula. Applications to spaces of real sequence-valued continuous functions give an explicit oscillation formula and exhibit the compatibility obstruction on an unbounded interval. Classical fixed-point consequences are recorded briefly.

fields
Functional Analysis
pdf ↗ page arXiv:2005.01832 ↗

This is a preliminary version of a book in progress on the theory of quantum communication. We adopt an information-theoretic perspective throughout and give a comprehensive account of fundamental results in quantum communication theory from the past decade (and earlier), with an emphasis on the modern one-shot-to-asymptotic approach that underlies much of today's state-of-the-art research in this field. In Part I, we cover mathematical preliminaries and provide a detailed study of quantum mechanics from an information-theoretic perspective. We also provide an extensive and thorough review of quantum entropies, and we devote an entire chapter to the study of entanglement measures. Equipped with these essential tools, in Part II we study classical communication (with and without entanglement assistance), entanglement distillation, quantum communication, secret key distillation, and private communication. In Part III, we cover the latest developments in feedback-assisted communication tasks, such as quantum and classical feedback-assisted communication, LOCC-assisted quantum communication, and secret key agreement.

comments
v3: 1252 pages; new co-author added; comments welcome!
fields
Information Theory, Mathematical Physics
pdf ↗ page arXiv:2011.04672 ↗

We show that for each ${\mathbf c}_{\mathrm M} \in [1,25)$, there is a unique metric associated with Liouville quantum gravity (LQG) with matter central charge ${\mathbf c}_{\mathrm M}$. An earlier series of works by Ding-Dubédat-Dunlap-Falconet, Gwynne-Miller, and others showed that such a metric exists and is unique in the subcritical case ${\mathbf c}_{\mathrm M} \in (-\infty,1)$, which corresponds to coupling constant $γ\in (0,2)$. The critical case ${\mathbf c}_{\mathrm M} = 1$ corresponds to $γ=2$ and the supercritical case ${\mathbf c}_{\mathrm M} \in (1,25)$ corresponds to $γ\in \mathbb C$ with $|γ| = 2$. Our metric is constructed as the limit of an approximation procedure called Liouville first passage percolation, which was previously shown to be tight for $\mathbf c_{\mathrm M} \in [1,25)$ by Ding and Gwynne (2020). In this paper, we show that the subsequential limit is uniquely characterized by a natural list of axioms. This extends the characterization of the LQG metric proven by Gwynne and Miller (2019) for $\mathbf c_{\mathrm M} \in (-\infty,1)$ to the full parameter range $\mathbf c_{\mathrm M} \in (-\infty,25)$. Our argument is substantially different from the proof of the characterization of the LQG metric for $\mathbf c_{\mathrm M} \in (-\infty,1)$. In particular, the core part of the argument is simpler and does not use confluence of geodesics.

code
github.com/leobon12/lean-formalizations/tree/main/results/supercritical-lqg-uniqueness ∎ Lean project (lakefile.lean, lean-toolchain) is too large to build within the check's time limit at 31e7483, checked 2026-11-06 build details
comments
109 pages, 24 figures; published in Proceedings of the London Mathematical Society; Lean formalization available at https://github.com/leobon12/lean-formalizations/tree/main/results/supercritical-lqg-uniqueness
journal
Prob. Math. Phys. 5 (2024) 1-54
fields
Probability, Mathematical Physics, Metric Geometry
pdf ↗ page arXiv:2110.00177 ↗

We establish the functional relations between generating series of higher-order free cumulants and moments in higher-order free probability, solving an open problem posed fifteen years ago by Collins, Mingo, Śniady and Speicher. We propose an extension of free probability theory, which governs the all-order topological expansion in unitarily invariant matrix ensembles, with a corresponding notion of free cumulants and give as well their relation to moments via functional relations. Our approach is based on the study of a master transformation involving double monotone Hurwitz numbers via semi-infinite wedge techniques, building on the recent advances of the last-named author with Bychkov, Dunin-Barkowski and Kazarian. We illustrate our formulas by computing the first few decaying terms of the correlation functions of an ensemble of spiked GUE matrices, going beyond the law of large numbers and the central limit theorem.

comments
93 pages. v3: some clarifications and corrections, to appear in Astérisque
fields
Operator Algebras, Mathematical Physics, Combinatorics, Probability
pdf ↗ page arXiv:2112.12184 ↗

We investigate singularly perturbed nonlinear complex differential systems of the form $\hbar \partial_x f = F (x, \hbar, f)$, where $\hbar$ is a small complex perturbation parameter. We give geometric and analytic conditions under which the formal perturbative solution associated with a fixed regular leading-order branch is locally uniformly Borel summable. This leads to an existence and uniqueness theorem of holomorphic solutions with prescribed perturbative expansion. We give an explicit triangular example and a self-contained application to the singularly perturbed Painlevé I equation.

comments
18 pages. Substantial revision with a new title
fields
Classical Analysis and ODEs, Mathematical Physics, Analysis of PDEs
pdf ↗ page arXiv:2201.04526 ↗

In this version of the paper, we note important gaps in the previous version.

fields
Representation Theory, Analysis of PDEs
pdf ↗ page arXiv:2304.07009 ↗

Let $β>1$ be fixed. We consider the $(\mathbf{b},\mathcal{D})$-numeration system, where the base sequence $\mathbf{b}=(b_k)_{k\geq 0}$ consists of positive real numbers satisfying $\lim_{k\rightarrow \infty}b_{k+1}/b_k=β$, and the digit set $\mathcal{D}\ni 0$ is a finite set of nonnegative real numbers containing at least two elements. Let $r_{\mathbf{b},\mathcal{D}}(λ)$ denote the number of representations of a given $λ\in\mathbb{R}$ in the form $\sum_{k\ge 0}δ_kb_k$ with digits $δ_k\in \mathcal{D}$. We establish upper bounds for $r_{\mathbf{b},\mathcal{D}}(λ)$ as well as asymptotic formulas for its arbitrary moments. Furthermore, we prove that the associated Dirichlet series $ζ_{\mathbf{b},\mathcal{D}}(s):=\sum_{λ>0}r_{\mathbf{b},\mathcal{D}}(λ)λ^{-s}$ admits a meromorphic continuation to the entire complex plane when $b_k=β^{k}$, and to the half-plane $\mathrm{Re}(s)>\log_β\mathcal{D}-γ$ when $b_k=β^{k}+O(β^{(1-γ)k})$ for any fixed $γ\in(0,1]$. Within the regions of continuation, we determine all candidate poles, compute their corresponding residues, and locate the trivial zeros of $ζ_{\mathbf{b},\mathcal{D}}(s)$. As an application, we resolve several problems posed by Chow and Slattery concerning partitions into distinct terms of certain integer sequences.

fields
Number Theory, Combinatorics
pdf ↗ page arXiv:2305.00792 ↗

We study the categorical-algebraic condition that internal actions are weakly representable (WRA) in the context of varieties of (non-associative) algebras over a field. Our first aim is to give a complete characterization of action accessible, operadic quadratic varieties of non-associative algebras which satisfy an identity of degree two and to study the representability of actions for them. Here we prove that the varieties of two-step nilpotent (anti-)commutative algebras are weakly action representable, and we explain that the condition (WRA) is closely connected to the existence of a so-called strong amalgam. Our second aim is to work towards the construction, still within the context of algebras over a field, of a weakly representing object $\mathcal{E}(X)$ for the actions on (or split extensions of) an object $X$. We actually obtain a partial algebra $\mathcal{E}(X)$, which we call external weak actor of $X$, together with a monomorphism of functors ${\mathrm{SplExt}(-,X) \rightarrowtail \mathrm{Hom}(U(-),\mathcal{E}(X))}$, which we study in detail in the case of quadratic varieties. Furthermore, the relations between the construction of the universal strict general actor $USGA(X)$ and that of $\mathcal{E}(X)$ are described in detail. We end with some open questions. This version is preceded by a corrigendum. Theorem 2.11 of the published article, which states that the variety of commutative associative algebras is weakly action representable, is false: we exhibit a $16$-dimensional algebra $X$ with zero multiplication for which the functor $\mathrm{SplExt}(-,X)$ admits no weak representation. Hence this variety is action accessible but not weakly action representable. The corrigendum amends the statements affected by this error, corrects Proposition 2.8, whose "if" part needs a strong amalgam, updates the definition of partial algebra, and lists some further minor corrections.

comments
v4: 8-page corrigendum prepended to the published version. Theorem 2.11 (weak action representability of commutative associative algebras) is false; a counterexample is given and the affected statements are amended. The remaining results are unaffected, apart from minor corrections listed at the end
journal
Journal of Algebra 669 (2025), pp. 401-444
fields
Category Theory, Rings and Algebras
pdf ↗ page arXiv:2306.02812 ↗

We develop a distributionally robust formulation of the single-airport ground holding problem (dr-SAGHP) using a $1$-Wasserstein ambiguity set with an $\ell_1$ transportation metric over the full physical capacity support. We exploit the problem's structural properties to develop an exact dynamic programming separation algorithm within delayed constraint generation (DCG-DP). We also develop K-IL, which combines Kelley's cutting-plane method with an integer L-shaped framework to solve a a restricted-support approximation, rs-dr-SAGHP. Delayed constraint generation with MILP separation (DCG-MILP) serves as the computational baseline. Computational experiments demonstrate substantial runtime reductions for DCG-DP and K-IL relative to this baseline. Airport capacity distributions are constructed using Gaussian process regression based on historical weather and operational data at Denver International Airport from 2021 to 2023. We develop a chronological expanding-window calibration procedure to select the Wasserstein radius and the results show that the calibrated dr-SAGHP approaches reduce mean operational cost by approximately $2.6\%$ and the Conditional Value at Risk $\mathrm{CVaR}_{90}$ by approximately $8\%$ relative to the stochastic SAGHP benchmark, while avoiding the excessive conservatism observed in a classical robust optimization benchmark.

comments
44 pages, 12 figures
fields
Optimization and Control
pdf ↗ page arXiv:2306.09836 ↗

We introduce a path metric $ρ_D$ on the shift locus $\mathcal{S}_D$ of degree $D\geq2$ complex polynomials by connecting complex dynamics to geometric group theory. We encode the escape and deformation data of polynomial basins by metric graphs and use the thermodynamic geometry of the resulting metric graph space. We prove that $(\mathcal{S}_D,ρ_D)$ is complete if and only if $D=2$. For $D\geq3$, we show that the positive-height part of the metric boundary is naturally homeomorphic to the projectivized DeMarco–Pilgrim space $\mathbb{P}\mathcal{S}\mathcal{T}_D^*$. Our result provides a new way of understanding this space which encodes the topology of the shift locus and is defined abstractly as a monotone-light quotient.

comments
29 pages
fields
Dynamical Systems
pdf ↗ page arXiv:2311.11399 ↗

A plurisubharmonic weight is \emph{log canonical} if it is at the critical point of turning non-integrable. Given a log canonical plurisubharmonic weight, we show that locally there always exists a log canonical `holomorphic' weight having the same non-integrable locus. The proof uses the theory of quasimonomial valuations developed by Jonsson-Mustaţă and Xu, as well as the minimal model program over complex analytic spaces due to Fujino and Lyu-Murayama. As a consequence, we obtain that the non-integrable locus is seminormal.

comments
Minor changes, to appear in Annales de l'Institut Fourier
fields
Complex Variables, Algebraic Geometry
pdf ↗ page arXiv:2312.16140 ↗

We study two variants of the mirror descent-ascent (MDA) algorithm for solving min-max problems on the space of measures: simultaneous and alternating. We work under assumptions of convexity-concavity and relative smoothness of the payoff function with respect to a suitable Bregman divergence, defined on the space of measures via flat derivatives. We establish non-asymptotic convergence rates to mixed Nash equilibria, measured in the Nikaidô-Isoda error, proving an $\mathcal{O}(N^{-1/2})$ rate for simultaneous MDA and an improved $\mathcal{O}(N^{-2/3})$ rate for alternating MDA. The main technical contribution is an infinite-dimensional dual space analysis that relates Bregman divergences on measures to dual Bregman divergences on spaces of bounded continuous functions, allowing us to control asymmetric commutator terms created by alternating updates. The results substantially generalize prior analyses restricted to bilinear objectives and also apply to nonlinear convex-concave problems on measure spaces, thereby providing a unified theoretical foundation for MDA in mean-field min-max optimization.

comments
Accepted at NeurIPS 2026
fields
Optimization and Control, Probability
pdf ↗ page arXiv:2402.08106 ↗

We consider the colored asymmetric simple exclusion process (ASEP) and stochastic six vertex (S6V) model with fully packed initial conditions; the states of these models can be encoded by 2-parameter height functions. We show under Kardar-Parisi-Zhang (KPZ) scaling of time, space, and fluctuations that these height functions converge to the Airy sheet. Several corollaries follow. (1) For ASEP and the S6V model under the basic coupling, we consider the 4-parameter height function at position $y$ and time $t$ with a step initial condition at position $x$ and time $s < t$, and prove that under KPZ scaling it converges to the directed landscape. (2) As a consequence, in our paper arXiv:2412.18117, we prove that ASEPs and S6V models under the basic coupling, with multiple general initial data, converge to KPZ fixed points coupled through the directed landscape. (3) We prove that the colored ASEP stationary measures converge to the stationary horizon. (4) We prove a form of decoupling for the colored ASEP height functions, as well as for the stationary two-point function, as broadly predicted by the theory of non-linear fluctuating hydrodynamics. The starting point for our Airy sheet convergence result is an embedding of these colored models into a larger structure -- a color-indexed family of coupled line ensembles with an explicit Gibbs property, i.e., a colored Hall-Littlewood line ensemble. The core of our work then becomes to develop a framework to analyze the edge scaling limit of these ensembles.

comments
148 pages, 28 figures. Corrected some typos and minor errors and expanded on a few proofs
fields
Probability, Mathematical Physics
pdf ↗ page arXiv:2403.01341 ↗

We revise a proof of a Mazur-Ulam theorem for generalized gyrovector spaces.

comments
13 pages
fields
Functional Analysis
pdf ↗ page arXiv:2403.15015 ↗

Let $Ω$ be an arbitrary domain in ${\bf R}^n$, and let $C^2(Ω)$ be the space of all functions whose partial derivatives of all orders up to two are continuous and bounded on $Ω$. We prove the following Finiteness Principle for the boundary values of $C^2(Ω)$-functions: A function $f$ defined on the boundary of $Ω$ is the trace of a function $F\in C^2(Ω)$ provided there exists $λ>0$ such that, for every set $E$ of at most $N=3\cdot 2^{n-1}$ boundary points, there exists a function $F_E\in C^2(Ω)$ with $\|F_E\|_{C^2(Ω)}\leλ$ whose trace coincides with $f$ on $E$. Moreover, $F$ can be chosen with $\|F\|_{C^2(Ω)}\leγλ$ where $γ$ depends only on $n$. We also refine this principle by showing that it suffices to consider $N$-point sets $E$ with specific geometric ``visibility'' properties relative to $Ω$.

comments
51 pages, 2 figures. This is the journal version of the paper
fields
Functional Analysis
pdf ↗ page arXiv:2406.04431 ↗

We propose a robust Q-learning algorithm for Markov decision processes under model uncertainty when each state–action pair is associated with a finite ambiguity set of candidate transition kernels. This finite-measure framework enables user-designed uncertainty models, including competing fitted models and stress scenarios. {The learner observes successor states and their deterministic rewards from every candidate simulator, without requiring explicit knowledge of the transition probabilities or the reward function.} We establish almost sure convergence of the learned Q-function to the robust optimum, and derive a time-uniform high-probability error certificate that separates an observable empirical Bellman residual from transition-kernel estimation error. We also give quantitative finite-ambiguity approximation results for Wasserstein and parametric uncertainty models on finite state spaces. Numerical examples illustrate the effect of scenario selection and compare the tabular algorithm with robust value iteration.

fields
Optimization and Control, Probability
pdf ↗ page arXiv:2407.04259 ↗

We propose a scalable approximate algorithm for the NP-hard maximum-weight independent set problem, based on dual coordinate descent applied to a smoothed clique-cover LP relaxation. Our method, a variant of the Bregman/Sinkhorn algorithm, employs entropy smoothing with a novel duality-gap-based smoothing scheduling strategy that empirically outperforms standard feasibility scheduling for the relaxed problem. Our new projection to the primal feasible set enables accurate duality gap estimation. Combined with a basic primal heuristic leveraging reduced costs, our approach yields high-quality integer solutions. On real-world datasets, it efficiently finds high-quality approximate solutions for graphs with up to 882,000 nodes and 344 million edges within seconds.

fields
Optimization and Control
pdf ↗ page arXiv:2408.02086 ↗

We study the comparison principle for viscosity sub- and super-solutions of degenerate quasi-linear equations in non-divergence form driven by the horizontal vector fields of a Carnot group, under an ellipticity condition and a homogeneity condition.

comments
Major revisions have been made
fields
Analysis of PDEs
pdf ↗ page arXiv:2409.15144 ↗

We derive necessary and sufficient conditions for universality limits for orthogonal polynomials associated to a fixed measure on the real line and related systems. One of our results is that the Christoffel–Darboux kernel has sine kernel asymptotics at a point $ξ$, with regularly varying scaling, if and only if the orthogonality measure (spectral measure) has a unique tangent measure at $ξ$ and that tangent measure is the Lebesgue measure. All previously known cases of sine kernel asymptotics, with absolutely continuous or singular measures, have regularly varying scaling. Our work is not limited to bulk universality; we show that the Christoffel–Darboux kernel has a regularly varying scaling limit with a nontrivial limit kernel if and only if the orthogonality measure has a unique tangent measure at $ξ$ and that tangent measure is not a point mass. The possible limit kernels correspond to homogeneous de Branges spaces; in particular, this equivalence completely characterizes several prominent universality classes such as hard edge universality, Fisher–Hartwig singularities, and jump discontinuities in the weights. A key ingredient in the proof is a new correspondence between measures and chains of de Branges spaces, not between Weyl functions and Hamiltonians. In order to handle limits with power law weights, this correspondence goes beyond the more common setting of Poisson-finite measures, and allows arbitrary power bounded measures.

comments
82 pages; to appear in Inventiones Mathematicae
fields
Classical Analysis and ODEs, Mathematical Physics, Complex Variables, Spectral Theory
pdf ↗ page arXiv:2409.18045 ↗

In this paper, we establish spherical maximal inequalities associated with commuting trace-preserving automorphisms on von Neumann algebras, extending Magyar-Stein-Wainger's discrete spherical maximal inequalities to the noncommutative setting. On a noncommutative probability space, we further prove that, under a strong ergodicity assumption, the corresponding discrete spherical averages converge bilaterally almost uniformly to $τ(x)1$ for every $x\in L_p(\mathcal M)$, $\frac{d}{d-2}<p\leq\infty$, when $d\geq5$.

comments
24 pages
fields
Operator Algebras, Functional Analysis
pdf ↗ page arXiv:2410.06035 ↗

It is well-known that conditioning a supercritical (multi-type) branching process on the event that it eventually becomes extinct yields a subcritical branching process. We study the corresponding inverse problem: given a subcritical branching process, does there exist a supercritical branching process with the property that when we condition it on extinction, we get back the original subcritical branching process? We show that such a supercritical branching process (which we call a conjugate branching process) exists under mild hypotheses on the original subcritical branching process. We also show by example that if there are at least two types, then the conjugate branching process is not necessarily unique. Our results are relevant to the problem of constructing natural random planar maps whose scaling limit is given by supercritical Liouville quantum gravity. Moreover, conjugate branching processes can also be used to give alternative evolutionary hypotheses in cancer modeling.

code
github.com/ewaingwynne/conjugate-branching-lean ∎ Lean project (lakefile.toml, lean-toolchain) builds, no sorry at 878bd03, checked 2026-11-06 build details
comments
29 pages, 3 figures; Lean formalization available at https://github.com/ewaingwynne/conjugate-branching-lean
journal
Electron. J. Probab. 30: 1-27 (2025)
fields
Probability
pdf ↗ page arXiv:2411.06301 ↗

Popa and Schnell show that any holomorphic 1-form on a smooth projective variety of general type has zeros. In this article, we show that a smooth good minimal model has a holomorphic 1-form without zero if and only if it admits an analytic fiber bundle structure over a positive dimensional abelian variety.

comments
Almost the same result has appeared in an earlier preprint arXiv:2410.22753, but the approach is different. To appear in Ann. Sc. Norm. Super. Pisa Cl. Sci. (5), DOI: https://doi.org/10.2422/2036-2145.202501_007
fields
Algebraic Geometry
pdf ↗ page arXiv:2412.12582 ↗

We show the existence and uniqueness of solutions to a generalized Monge-Ampère equation on closed almost Kähler surfaces, where the equation depends only on the underlying almost Kähler structure. As an application, we prove Donaldson's conjecture for tamed almost complex 4-manifolds.

comments
Theorem 1.1 fixed
fields
Differential Geometry, Analysis of PDEs
pdf ↗ page arXiv:2412.18361 ↗

We calculate the automorphism group of the Kummer surface associated with a curve of genus 2 or the product of two elliptic curves in characteristic two under the assumption that the Kummer surface is a $K3$ surface. Moreover we discuss the complex $K3$ surfaces with the same Picard lattice as these Kummer surfaces. The paper has two appendices.

comments
Final version; 48 pages, 23 Figures
fields
Algebraic Geometry
pdf ↗ page arXiv:2412.18902 ↗

We establish new connections between multitype $Λ$-coalescents and multitype continuous state branching processes (CSBPs) through duality with a $Λ-$Wright-Fisher frequency process, and a homeomorphism between the spaces of the two classes of processes. The approach relies on a sequential sampling procedure for the frequency process of multitype CSBPs and provides an interpretation of multitype $Λ-$coalescents as the local genealogies of multitype CSBPs which gives further insight into different approaches to generalise $Λ-$coalescents to the multitype setup. As an application of the homeomorphism, we introduce and discuss multitype Beta-coalescents as the local genealogy of multitype $α$-stable branching processes.

fields
Probability
pdf ↗ page arXiv:2501.00329 ↗

Let $K$ be a complete discretely valued field of dimension $2$. We prove the following statements on the arithmetic of $K$. - The ``period equals index'' property holds for central simple $K$-algebras. - If $K$ has characteristic $0$, for every prime $p$, every class in the Milnor $\mathrm{K}$-theory modulo $p$ is represented by a symbol. - Serre's Conjecture II holds for the field $K$. That is, for every semisimple and simply connected $K$-group $G$, the set $H^1(K,G)$ is trivial.

comments
31 pages. Accepted for publication. Includes several improvements suggested by an anonymous referee
fields
Rings and Algebras, Algebraic Geometry, K-Theory and Homology
pdf ↗ page arXiv:2501.01403 ↗

Let $M$ be a compact torsion-free $G_2$ 7-manifold or Calabi-Yau 6-manifold. We prove Hodge decomposition theorems for the $dd^φ$ operators, introduced by Harvey and Lawson, which generalize the $i\partial\bar\partial$ operator used in classical pluripotential theory. We then obtain analogues of the $\partial\bar\partial$ lemma in this context. We formalize this by defining cohomology spaces analogous to Bott-Chern cohomology and we relate them to harmonic forms on $M$. In the $G_2$ case we provide a geometric interpretation of the corresponding cohomology classes in terms of coassociative submanifolds and gerbes: this is analogous to the classical interpretation of Bott-Chern cohomology classes in terms of divisors and holomorphic line bundles.

comments
v3: final version
journal
The Journal of Geometric Analysis 36, 393 (2026)
fields
Differential Geometry
pdf ↗ page arXiv:2501.03778 ↗

We prove absence of ground states in the infrared-divergent spin boson model at large coupling. Our key argument reduces the proof to verifying long range order in the dual one-dimensional continuum Ising model, i.e., to showing that the respective two point function is lower bounded by a strictly positive constant. We can then use known results from percolation theory to establish long range order at large coupling. Combined with the known existence of ground states at small coupling, our result proves that the spin boson model undergoes a phase transition with respect to the coupling strength. We also present an expansion for the vacuum overlap of the spin boson ground state in terms of the Ising $n$-point functions, which implies that the phase transition is unique, i.e., that there is a critical coupling constant below which a ground state exists and above which none can exist.

comments
19 pages, version accepted for publication in Forum of Mathematics, Sigma
fields
Mathematical Physics
pdf ↗ page arXiv:2501.19362 ↗

The stochastic Airy and sine operators, which are respectively a random Sturm-Liouville operator and a random Dirac operator, characterize the soft edge and bulk scaling limits of $β$-ensembles. Dirac and Sturm-Liouville operators are distinct operator classes which can both be represented as canonical systems, which gives a unified framework for defining important properties, such as their spectral data. Seeing both as canonical systems, we prove that in a suitable high-energy scaling limit, the Airy operator converges to the sine operator. We prove this convergence in the vague topology of canonical systems' coefficient matrices, and deduce the convergence of the associated Weyl-Titchmarsh functions and spectral measures. Our proof relies on a coupling between the Brownian paths that drive the two operators, under which the convergence holds in probability. This extends the corresponding result at the level of the eigenvalue point processes, proven by Valkó and Virág (2009) by comparison to the Gaussian $β$-ensemble.

comments
61 pages
fields
Probability, Mathematical Physics, Spectral Theory
pdf ↗ page arXiv:2502.10305 ↗

A graph is rigid if it only admits the identity endomorphism. We show that for every $d\ge 3$ there exist infinitely many mutually rigid $d$-regular graphs of arbitrary odd girth $g\geq 7$. Moreover, we determine the minimum order of a rigid $d$-regular graph for every $d\ge 3$. This provides strong positive answers to a question of van der Zypen [https://mathoverflow.net/q/296483, https://mathoverflow.net/q/321108]. Further, we use our construction to show that every finite monoid is isomorphic to the endomorphism monoid of a regular graph. This solves a problem of Babai and Pultr [J. Comb. Theory, Ser. B, 1980].

comments
34 pages, 13 figures, journal (revised) version
journal
European Journal of Combinatorics 140 (2027) 104466
fields
Combinatorics
pdf ↗ page arXiv:2502.11421 ↗

This paper introduces a fractal $ψ$-Fueter operator in the quaternionic context inspired by the concepts of proportional derivative and Hausdorff derivative ({fractal derivative}) of a function with respect to a fractal measure. Moreover, we establish the corresponding Stokes and Borel-Pompeiu formulas associated to this generalized fractal $ψ$-Fueter operator with respect to a vector valued function induced by truncated exponential functions.

comments
35 pages, 0 figures
fields
Complex Variables
pdf ↗ page arXiv:2502.15518 ↗

In this note we prove that the Langevin Monte Carlo (LMC) algorithm can minimize the population risk of appropriately regularized depth-2 nets with no restrictions on the size of the net and for any choice of training data --- and we give non-asymptotic convergence rates for it. We achieve this via showing that in q-Renyi divergence, the iterates of LMC converge to the Gibbs distribution of certain Frobenius norm regularized losses for these nets --- in both classification and regression settings. The key technicality of the work is two-fold (a) to keep the amount of regularization needed be independent of the size of the net and (b) to obtain convergence rates that scale with just the Poincare constant of the corresponding Gibbs measure. This work synthesizes several recent observations about isoperimetry conditions under which LMC converges and that two-layer neural loss functions can always be regularized by a constant amount such that they satisfy the Villani conditions, and thus their Gibbs measures satisfy a Poincare inequality.

fields
Functional Analysis, Probability
pdf ↗ page arXiv:2503.10428 ↗

The averaging principle for slow-fast systems of various kind of stochastic (partial) differential equations has been extensively studied. An analogous result was shown for slow-fast systems of rough differential equations driven by random rough paths a few years ago and the study of ``rough slow-fast systems" seems to be gaining momentum now. In all known results, however, the driving rough paths are of level 2. In this paper we formulate rough slow-fast systems driven by random rough paths of level 3 and prove the strong averaging principle of Khas'minskii-type.

comments
A minor error in the previous version (Lemma 5.9) was corrected. 46 pages. No figure
fields
Probability, Dynamical Systems
pdf ↗ page arXiv:2504.03110 ↗

Global permutation patterns have recently been shown to characterize important properties of a Coxeter group. Here we study global patterns in the context of signed permutations, with both characterizing and enumerative results. Surprisingly, many properties of signed permutations may be characterized by avoidance of the same set of patterns as the corresponding properties in the symmetric group. We also extend previous enumerative work of Egge, and our work has connections to the Garfinkle–Barbasch–Vogan correspondence, the Erdős–Szekeres theorem, and well-known integer sequences.

comments
22 pages. v2: major revision and expanded discussion of existing literature. v3: minor correction in Section 7.2
fields
Combinatorics
pdf ↗ page arXiv:2504.13108 ↗

In this paper, we study the following nonlinear Schrödinger system of Hamiltonian type \begin{equation*} \left\{\begin{array}{l} -Δu+V(x)u=\partial_v H(x,u,v)+ωv,\ x \in \mathbb{R}^N, \\ -Δv+V(x)v=\partial_u H(x,u,v)+ωu,\ x \in \mathbb{R}^N, \\ \displaystyle\int_{\R^N}|z|^2dx=a^2, \end{array}\right. \end{equation*} where the potential function $V(x)$ is periodic, $z:=(u,v):\R^N\rightarrow \R\times\R$, $ω\in \R$ appears as a Lagrange multiplier, $a>0$ is a prescribed constant. The existence and multiplicity of $L^2$-normalized solutions for the above Schrödinger system are obtained. The proof combines a constrained variational reduction, compactness estimates, and a Ljusternik–Schnirelmann type multiplicity theorem. In addition, a bifurcation result is also given.

comments
37 pages
fields
Analysis of PDEs
pdf ↗ page arXiv:2504.15656 ↗

In this paper, we generalize Zhou valuations, originally defined on complex domains, to the framework of general schemes. We demonstrate that an algebraic version of the Jonsson–Mustaţă conjecture is equivalent to the statement that every Zhou valuation is quasi-monomial. By introducing a mixed version of jumping numbers and Tian functions associated with valuations, we obtain characterizations of a valuation being a Zhou valuation or computing some jumping number using the Tian functions. Furthermore, we establish the correspondence between Zhou valuations in algebraic settings and their counterparts in analytic settings.

comments
We fixed some mistakes in the latest version
fields
Algebraic Geometry, Complex Variables
pdf ↗ page arXiv:2505.19451 ↗

In graph pebbling, pebbles are distributed across the vertices of a graph. A pebbling move consists of removing two pebbles from one vertex and adding one pebble to an adjacent vertex. The pebbling number is the minimum number $t$ such that, regardless of how the $t$ pebbles are initially placed, it is always possible to perform a sequence of moves that results in at least one pebble placed on any specified target vertex. Graphs whose pebbling number is equal to the number of vertices are called Class~$0$ and provide a challenging set of graphs that resist being characterized. In this note, we answer a question recently proposed by the pioneering study on the pebbling number of snark graphs: we prove that the smallest Flower snark $J_3$ is Class~$0$, establishing that $J_3$ in fact is the only Class~$0$ Flower snark. that systematically reduce the space of pebble configurations. Our bound-based approach shows particular promise for diameter $3$ graphs, where the neighborhood structure remains sufficiently constrained to enable effective analysis.

comments
10 pages, 6 figures. Proof reorganized and discussion improved. Supplementary software available at https://github.com/gabridi/pebbling_unsolvability
fields
Combinatorics
pdf ↗ page arXiv:2505.22941 ↗

Proximal Policy Optimization (PPO) is widely used in reinforcement learning due to its strong empirical performance, yet it lacks formal guarantees for policy improvement and convergence. PPO's clipped surrogate objective is motivated by a lower bound on linearization of the value function in flat geometry setting. We derive a tighter surrogate objective and introduce Fisher-Rao PPO (FR-PPO) by leveraging the Fisher-Rao (FR) geometry. Our scheme provides strong theoretical guarantees, including monotonic policy improvement. In the direct parametrization setting, we show that FR-PPO achieves sub-linear convergence, and for parametrized policies we further obtain sub-linear convergence up to the compatible function approximation error. Finally, although our primary focus is theoretical, we also demonstrate empirically that FR-PPO can be implemented as a drop-in replacement for PPO and achieves performance comparable to PPO across a range of standard reinforcement learning tasks.

comments
Accepted at NeurIPS 2026
fields
Optimization and Control
pdf ↗ page arXiv:2506.03757 ↗

Previously we found a unique quantum system with a positive gauge-invariant Weyl-Stratonovich quasi-probability density function which can be defined by the so-called «quadratic funnel» potential. In this work we have constructed a class of exact solutions to the 3D Schrödinger equation for a two-parameter «quadratic funnel» potential based on the $Ψ$ -model of micro and macro systems. It should be noted that the Aharonov-Bohm effect is inherent in the quantum $Ψ$ -model. Explicit expressions for the energy spectrum and the set of eigenfunctions have been found. Using gauge invariance for scalar and vector potentials, a solution to the electromagnetic Schrödinger equation has been obtained, with a magnetic field in the form of a «Dirac string» defined by a singular vortex probability flux field. Superpositions of eigenfunctions leading to various types of vortex and potential probability current fields have been investigated in detail. The analysis of the quantum system's properties has been carried out within the Wigner-Vlasov formalism.

comments
32 pages, 6 figures
fields
Mathematical Physics
pdf ↗ page arXiv:2506.04725 ↗

Building on ideas of Kohno, we develop a framework for the construction of higher holonomy functors via the transport of formal power series connections. Using these techniques, we obtain functors from the path groupoid, the path 2-groupoid, and the path 3-groupoid of a manifold. We show that the resulting higher holonomy functors can detect non-trivial classes in higher homotopy groups. As an application, we construct a Gray functor from the path 3-groupoid of the configuration space of $m$ points in $\mathbb{R}^n$ for $n\geqslant 4$ and compute its restriction to 3-tracks with constant boundary when $n=4$.

comments
22 pages. v2: added more examples, more general formulation of formal power series connections in terms of algebras instead of coalgebras, added details to final section, corrected typos
fields
Algebraic Topology, Category Theory, Quantum Algebra
pdf ↗ page arXiv:2506.23880 ↗

We consider the Cauchy problem for the modified Camassa-Holm equation \[ u_t+\left((u^2-u_x^2)m\right)_x=0,\quad m:= u-u_{xx}, \quad t>0,\ \ -\infty<x<+\infty \] subject to the step-like initial data: $u(x,0)\to A_1$ as $x\to-\infty$ and $u(x,0)\to A_2$ as $x\to+\infty$, where $0<A_1<A_2$. First, using the Riemann–Hilbert approach, we establish the global solvability of the corresponding problem in the reciprocal $(y,t)$-variables. Then the reconstruction in the original $(x,t)$-variables is obtained under a positivity condition expressed in terms of the solution of the associated Riemann–Hilbert problem.

comments
49 pages
fields
Analysis of PDEs
pdf ↗ page arXiv:2507.02100 ↗

Lin, Chan (High order entropy stable discontinuous Galerkin spectral element methods through subcell limiting, 2024) enforces a cell entropy inequality for nodal discontinuous Galerkin methods by combining flux corrected transport (FCT)-type limiting and a knapsack solver, which determines optimal limiting coefficients that result in a semi-discrete cell entropy inequality while preserving nodal bounds. In this work, we provide a slight modification of this approach, where we utilize a quadratic knapsack problem instead of a standard linear knapsack problem. We prove that this quadratic knapsack problem can be reduced to efficient scalar root-finding. Numerical results demonstrate that the proposed quadratic knapsack limiting strategy is efficient and results in a semi-discretization with improved regularity in time compared with linear knapsack limiting, while resulting in fewer adaptive timesteps in shock-type problems.

fields
Numerical Analysis
pdf ↗ page arXiv:2507.14488 ↗

We consider homothetic evolutions of the area-preserving curve-shortening flow (APCSF), that is, classical curve-shortening flow with an additional non-local forcing term. By using known results on $λ$-curves, we prove the existence of non-circular shrinkers for this flow. In our first main result, we present a full classification of such curves, similar to the well-known Abresch-Langer classification for shrinkers of curve-shortening flow. Finally, we also deduce a saddle-point property for all non-circular (APCSF)-shrinkers analogous to the known saddle-point property of Abresch-Langer curves.

fields
Differential Geometry
pdf ↗ page arXiv:2508.00352 ↗

We prove the existence of a finite connected flexible $3$-valent vertex-transitive graph of girth $\ell$ for every integer $\ell \geq 3$.

comments
11 pages -- Improved version. We extend the result from v1 to graphs of arbitrary girth
fields
Group Theory, Combinatorics
pdf ↗ page arXiv:2508.16289 ↗

We introduce a set membership perspective on gradient estimation for black-box scalar functions, in which the gradient is not approximated as a single point estimate but characterised as a member of a set of admissible gradients consistent with available function samples and regularity assumptions. Bounds on gradient uncertainty, obtained from standard multivariate Taylor series results, are used to represent this admissible set through linear inequalities, making gradient uncertainty explicit and directly exploitable. Building on this formulation, we propose an adaptive strategy to refine the admissible gradient set to a desired precision by selectively acquiring new samples only when necessary. The resulting framework naturally accommodates function evaluations affected by noise with finite but a priori unknown bounds, yields the theoretical best attainable precision of the gradient set, and provides the optimal sampling distance from the point of interest to maximise information gain during refinement. Within this framework, we introduce the Set-Based Adaptive Gradient Estimator (SAGE), which extracts an estimate of the true gradient from the admissible set whilst preserving high sample efficiency and robustness to noise. Numerical experiments on large collections of random gradient-query points, as well as in the context of a first-order optimisation method, compared the performance of SAGE against commonly used gradient estimators from the literature. Our results show that SAGE achieved competitive gradient accuracy across noise levels without requiring any a priori noise characterisation. In a first-order optimisation setting, noise-informed competitors can outperform SAGE when such characterisation is available. Nevertheless, SAGE remained broadly competitive, making it particularly attractive in settings where noise characterisation is impractical to obtain.

fields
Optimization and Control
pdf ↗ page arXiv:2508.19400 ↗

We introduce the notion of orthomesy, which is related to homomesy, a central phenomenon in dynamical algebraic combinatorics. An orbit is said to be orthomesic with respect to a statistic if the orbit's average statistic value is equal to the global average. We particularly focus on the action of promotion on increasing labelings of certain fence posets called zig-zag posets, and two statistics, the antipodal sum statistic and the total sum statistic. We classify all of the orthomesic promotion orbits for the zig-zag poset with four elements. Along the way, we investigate how homomesy of one action can be used to find orthomesic orbits for another action, for the same fixed statistic. We prove several general results which can be used to find infinite families of orthomesic orbits for any self-dual poset that has an order-reversing involution.

comments
Comments welcome! v2: we replaced the word "orbitmesy" with "orthomesy" and fixed typos
fields
Combinatorics
pdf ↗ page arXiv:2508.19440 ↗

To achieve the greatest possible speed, practitioners regularly implement randomized algorithms for low-rank approximation and least-squares regression with structured dimension reduction maps. Despite significant research effort, basic questions remain about the design and analysis of randomized linear algebra algorithms that employ structured random matrices. This paper develops a new perspective on structured dimension reduction, based on the oblivious subspace injection (OSI) property. The OSI property is a relatively weak assumption on a random matrix that holds when the matrix preserves the length of vectors on average and, with high probability, does not annihilate any vector in a low-dimensional subspace. With the OSI abstraction, the analysis of a randomized linear algebra algorithm factors into two parts: (i) proving that the algorithm works when implemented with an OSI; and (ii) proving that a given random matrix model has the OSI property. This paper develops both parts of the program. First, it analyzes standard randomized algorithms for low-rank approximation and least-squares regression under the OSI assumption. Second, it identifies many examples of OSIs, including random sparse matrices, randomized trigonometric transforms, and random matrices with tensor product structure. These theoretical results imply faster, near-optimal runtimes for several fundamental linear algebra tasks. The paper also provides guidance on implementation, along with empirical evidence that structured random matrices offer exemplary performance for a range of synthetic problems and contemporary scientific applications.

fields
Numerical Analysis
pdf ↗ page arXiv:2508.21189 ↗

Entropy stable methods have become increasingly popular in the field of computational fluid dynamics. They often work by satisfying some form of a discrete entropy inequality: a discrete form of the 2nd law of thermodynamics. Schemes which satisfy a (semi-)discrete entropy inequality typically behave much more robustly, and do so in a way that is hyperparameter free. Recently, a new strategy was introduced to construct entropy stable discontinuous Galerkin methods: knapsack limiting, which blends together a low order, positivity preserving, and entropy stable scheme with a high order accurate scheme, in order to produce a high order accurate, entropy stable, and positivity preserving scheme. Another recent strategy introduces an entropy correction artificial viscosity into a high order scheme, aiming to satisfy a cell entropy inequality. In this work, we introduce the techniques of knapsack limiting and artificial viscosity for finite difference discretizations. The proposed schemes preserve high order accuracy in sufficiently smooth conditions, are entropy stable, and are hyperparameter free. Moreover, the proposed knapsack limiting scheme provably preserves positivity for the compressible Euler and Navier-Stokes equations. Both schemes achieve this goal without significant performance tradeoffs compared to state of the art stabilized schemes.

comments
arXiv admin note: text overlap with arXiv:2507.14488
fields
Numerical Analysis
pdf ↗ page arXiv:2508.21226 ↗

In this paper and its companion [DEP2] we construct and study special generating sets for $p$-arithmetic unitary Lie groups, extending the seminal works of Lubotzky, Phillips and Sarnak to higher dimensions. The real embeddings of our sets form optimal topological generators for compact unitary Lie groups, and their $p$-adic embeddings give rise to Cayley graphs and complexes over finite unitary groups, which exhibit the cutoff phenomenon. In the current paper we prove two main results: First, we construct $p$-arithmetic lattices in $PU_4$ and $PU_8$ which act transitively on their corresponding Bruhat-Tits buildings. These groups are used to construct Golden Gate sets on two and three qubits, which have optimal circuit length and efficient compiling algorithms. Second, we study the cutoff phenomenon for quotients of the Bruhat-Tits building of a general semisimple $p$-adic group. We show that cutoff occurs for quotients which satisfy either the Ramanujan property or the weaker Sarnak-Xue Density Hypothesis (SXDH). For this purpose we bound the operator norms of a large class of Hecke operators in terms of the rate of decay of unitary representations. In the companion paper [DEP2], we prove such a SXDH result for unitary groups.

comments
The previous version has been split to two papers, this is the first part
fields
Number Theory, Representation Theory
pdf ↗ page arXiv:2509.09047 ↗

We consider Bernoulli$(p)$ percolation on (non)unimodular quasi-transitive graphs and prove that, almost surely, for any two clusters $C$ and $C'$, the set of vertices in $C$ within distance one of $C'$ is light, i.e.\ it has finite total modular weight. This is a significant step towards resolving a longstanding question of Häggström, Peres, and Schonmann, and a generalization of a theorem of Timár, who proved the same result in the unimodular setting. We present two proofs. The first one is a hyperfiniteness argument that is specific to Bernoulli$(p)$ percolation and is new even in the unimodular case; the second is more general and follows Timár's approach, but requires the development of weighted analogues of several classical unimodular results. The new insight in both proofs is nonhyperfiniteness, which we obtain from the recently developed theory of measure class preserving equivalence relations and graphs. In particular, along the way we establish a generalization of a theorem of Benjamini, Lyons, and Schramm on the existence of an invariant random subgraph with positive weighted Cheeger constant, which is of independent interest.

comments
24 pages, 3 figures. Added a new proof and improved presentation
fields
Probability, Dynamical Systems
pdf ↗ page arXiv:2509.10631 ↗

In this paper, we consider the problem of Gaussian approximation for the online linear regression task. We derive the corresponding rates for the setting of a constant stepsize and study the explicit dependence of the convergence rate on the problem dimension $d$ and quantities related to the design matrix. When the number of iterations $n$ is known in advance, our results yield a normal approximation rate of order $\sqrt{\log{n}/n}$, provided that the sample size $n$ is large enough.

comments
Published in Statistics (Taylor & Francis, 2026)
journal
Statistics (2026), 1-25
fields
Optimization and Control
pdf ↗ page arXiv:2509.14039 ↗

System identification in modern engineering systems faces emerging challenges from unanticipated adversarial attacks beyond existing detection mechanisms. In this work, we obtain a provably accurate estimate of the Markov parameter matrix of order $k$ to identify partially observed linear systems, in which the probability $p$ of having an attack at each time is smaller than $0.5$. We show that given the batch data accumulated up to time $T^*$, the $\ell_2$-norm estimator achieves an error with stepwise exponential decay as the attack probability $p$ decreases. Furthermore, low-dimensional attack structures can decouple the exponential decay order from the attack probability $p$, achieving an error decay that scales with $T^*$. We then propose a stochastic projected subgradient descent algorithm on streaming data that produces an estimate at each time $t<T^*$, in which case the expected estimation error proves to be the larger of $\mathcal{O}(k/\sqrt{t})$ and the baseline error of the $\ell_2$-norm estimator at $T^*$. This stochastic approach illustrates how non-smooth estimators can leverage first-order methods despite lacking recursive formulas. Finally, we integrate batch and streaming estimations to recover the Hankel matrix using the appropriate estimates of the Markov parameter matrix, which enables the synthesis of a robust adaptive controller based on the estimated balanced truncated model under adversarial attacks.

comments
16 pages, 2 figures
fields
Optimization and Control
pdf ↗ page arXiv:2509.15794 ↗

In the present paper we give very simple general statements which deal with approximation of a real number by rationals and are related to isolation phenomenon. In particular we study functions $ f(x)>f_1(x)>0$ such that existence of solutions $\frac{p}{q}$ of Diophantine inequality $ \left| α-\frac{p}{q}\right|< \frac{f(q)}{q^2} $ leads to the existence of solutions of inequality $ \left| α-\frac{p}{q}\right|< \frac{f_1(q)}{q^2} $.

fields
Number Theory
pdf ↗ page arXiv:2509.25628 ↗

The manifold hypothesis suggests that data lies on manifolds with smaller intrinsic dimension (ID) than their ambient dimension. However there is no empirical agreement on the estimates for ID from different estimators for realistic datasets. Thus it is important to test ID estimators (IDEs) with targeted stressors. In this work, we consider the role of anisotropy. To this end, we propose HomID, a collection of homogeneous spaces with anisotropic embedding, for benchmarking ID estimators. We observe that methods that perform well on standard benchmarks systematically degrade on HomID under identical resource allocation. We further observe that anisotropic distortion of such benchmarks also results in performance degradation. Finally, we demonstrate that controlled anisotropic distortions induce systematic shifts in the distributions on which these methods rely, providing a concrete mechanism for the resulting estimation errors in two particular IDEs. The code for the project is available at https://github.com/a-writes-code/homid

comments
17 figures, 37 pages. Accepted for publication in Transactions on Machine Learning Research (TMLR). OpenReview: https://openreview.net/forum?id=aBJAPDsNvm
fields
Metric Geometry
pdf ↗ page arXiv:2510.01335 ↗

Given a finite-dimensional complex simple Lie algebra $\mathfrak{g}$ and a complex number $κ$, let $V^κ(\mathfrak{g})$ be the associated universal affine vertex algebra. Gorelik and Kac [GK07] gave a sufficient and necessary condition for $V^κ(\mathfrak{g})$ to be simple. In this paper, for simply-laced $\mathfrak{g}$ and non-critical $κ$, we determine the weights of singular vectors of $V^κ(\mathfrak{g})$ with minimal conformal weights, when $V^κ(\mathfrak{g})$ is not simple. We further determine all the longest Weyl elements in the Kashiwara-Tanisaki character theorem [KT00] which correspond to the weights of the singular vectors.

comments
Corrected title metadata and added author affiliations and email addresses. Mathematical content unchanged. 53 pages
fields
Quantum Algebra
pdf ↗ page arXiv:2510.04445 ↗

In many applications, one seeks to approximate integration against a positive measure of interest by a positive discrete measure: a numerical quadrature rule with positive weights. One common desired discretization property is moment preservation over a finite dimensional function space, e.g., bounded-degree polynomials. Carathéodory's theorem asserts that if there is any finitely supported quadrature rule with more nodes than the dimension of the given function space, one can form a smaller (and hence more efficient) positive, nested, quadrature rule that preserves the moments of the original rule. We describe an efficient streaming procedure for Carathéodory-Steinitz pruning, a numerical procedure that implements Carathéodory's theorem for this measure compression. The new algorithm makes use of Givens rotations and on-demand storage of arrays to successfully prune very large rules whose storage complexity only depends on the dimension of the function space. This approach improves on a naive implementation of Carathéodory-Steinitz pruning whose runtime and storage complexity are quadratic and linear, respectively, in the size of the original measure. We additionally prove mathematical stability properties of our method with respect to a set of admissible, total-variation perturbations of the original measure. Our method is compared to two alternative approaches with larger storage requirements: non-negative least squares and linear programming, and we demonstrate comparable runtimes, with improved stability and storage robustness. Finally, we demonstrate practical usage of this algorithm to generate quadrature for discontinuous Galerkin finite element simulations on cut-cell meshes.

comments
Minor text updates, runtime comparison figure regenerated correcting for scaling constant, inclusion of python API. Fixes of minor technical errors in appendix and supplemental proofs
fields
Numerical Analysis
pdf ↗ page arXiv:2510.14916 ↗

We provide a rigorous random matrix theory analysis of spiked cross-covariance models where the signals across two high-dimensional data channels are partially aligned. These models are motivated by multi-modal learning and form the standard generative setting underlying Partial Least Squares (PLS), a widely used yet theoretically underdeveloped method. We show that the leading singular values of the sample cross-covariance matrix undergo a Baik-Ben Arous-Peche (BBP)-type phase transition, and we characterize the precise thresholds for the emergence of informative components. Our results yield the first sharp asymptotic description of the signal recovery capabilities of PLS in this setting, revealing a fundamental performance gap between PLS and the Bayes-optimal estimator. In particular, we identify the SNR and correlation regimes where PLS fails to recover any signal, despite detectability being possible in principle. These findings clarify the theoretical limits of PLS and provide guidance for the design of reliable multi-modal inference methods in high dimensions.

comments
25 pages, 7 figures
journal
AISTATS 2026
fields
Statistics Theory
pdf ↗ page arXiv:2510.17561 ↗

Based on the multidimensional Leibniz integral rule, we propose a novel unbiased stochastic derivative estimation framework for sample performance functions that are discontinuous in the parameter of interest. In this framework, for general discontinuities, a change of variables followed by the Leibniz integral rule generally yields a multi-run estimator, whereas for indicator-type discontinuities, a suitable transformation of the sample performance leads directly to a single-run estimator. The new framework includes existing likelihood ratio (LR)-based approaches as special cases and applies more broadly. In particular, it encompasses generalized LR (GLR) methods and the new formulation permits us to obtain much milder and simpler sufficient regularity conditions for these estimators to be unbiased, compared with the conditions given in previous papers. We also provide guidance on how to construct efficient single-run estimators in settings where existing methods may require the simulation of multiple sample paths or may be inapplicable. We validate the effectiveness and robustness of the proposed framework through an inventory-price control problem.

fields
Probability
pdf ↗ page arXiv:2511.00006 ↗

Mixed-integer nonlinear optimization (MINLP) comprises a large class of problems that are challenging to solve and exhibit a wide range of structures. The Boscia framework (Hendrych et al., 2025b) focuses on convex MINLP where the nonlinearity appears in the objective only. This paper provides an overview of the framework, showcases extensions post-publication and practical examples to illustrate its use and customizability. One key aspect is the integration and exploitation of Frank-Wolfe methods as continuous solvers within a branch-and-bound framework, enabling inexact node processing, warm-starting and explicit use of combinatorial structure among others. Three examples illustrate its flexibility, the user control over the optimization process and the benefit of oracle-based access to the objective and its gradient. Additionally, ablation studies are performed on the three examples to investigate the performance impact of the different features and customizations. The aim of this tutorial is to provide readers with an understanding of the main principles of the framework.

comments
42 pages, 11 figures. 8 tables
fields
Optimization and Control
pdf ↗ page arXiv:2511.01479 ↗

Generalized linear models (GLMs) are fundamental tools for statistical modeling, with maximum likelihood estimation (MLE) serving as the classical approach for parameter inference. While MLE performs well for canonical GLMs, it can become computationally challenging in more general settings with non-canonical, non-smooth, or saturating link functions, where the resulting optimization landscape may be ill-conditioned, non-convex, or non-differentiable. In this paper, we study an alternative estimation framework based on variational inequalities (VIs), which formulates GLM estimation through an operator-based equilibrium condition rather than likelihood minimization. We analyze the VI estimator from a statistical perspective and establish finite-sample error bounds and asymptotic normality, together with convergence guarantees for fixed-point and stochastic approximation algorithms. The framework accommodates a broad class of link functions, including non-canonical, non-smooth, and non-monotone cases satisfying a strong Minty-type condition, and extends naturally to generalized additive models (GAMs) via basis expansion. Numerical experiments demonstrate that under a fixed iteration budget, the VI approach achieves competitive estimation accuracy and faster, more stable convergence than MLE, particularly in GLMs and GAMs with non-canonical or non-smooth link functions.

comments
NeurIPS 2026
fields
Optimization and Control, Statistics Theory
pdf ↗ page arXiv:2511.03087 ↗

Given a compact space $X$ that does not admit an embedding (an injective continuous function) into $\mathbb{R}^d$, we study the discontinuity that any injective function $X \to \mathbb{R}^d$ must have. To this end, we define a scale invariant modulus of discontinuity and obtain general lower bounds, thus obtaining quantified nonembeddability results of Haefliger–Weber type. Moreover, we establish analogous lower bounds for simplicial complexes that do not admit an almost $r$-embedding in $\mathbb{R}^d$, thus obtaining a quantified version of the topological Tverberg theorem.

comments
18 pages, 3 figures. Version accepted for publication; updated author list: Sunhyuk Lim is no longer an author by mutual agreement
journal
Journal of Applied and Computational Topology 10, 31 (2026)
fields
Metric Geometry, Geometric Topology
pdf ↗ page arXiv:2511.07636 ↗

We develop a fully constructive, conservative, and collision-level realization of Kac's program for the spatially homogeneous Landau equation across the full interaction range, including the Coulomb case. Our model is the microcanonical binary-collision (MBC) process: a reversible pure-jump $N$-particle Markov process, realizing the grazing-collision mechanism via small conservative rotations of relative velocities. Under suitable assumptions on the initial data, we prove propagation of chaos in the joint mean-field and grazing-collision limit ($N\to\infty, h\to0$), identifying any limit point with the unique solution in the relevant weak-solution class. The empirical measures converge in quadratic Wasserstein distance, while fixed-order marginals converge strongly in $L^1$ at every fixed time. Furthermore, we rigorously derive the Landau master equation as the fixed-$N$ grazing-collision limit of the MBC process. The proof combines Fisher-information dissipation, compactness, and uniqueness; for $γ\in[-3,-2)$, second-order information along collision directions yields the time-integrable $L^\infty$ bound needed for uniqueness.

comments
This revision fixes the gap in the previous version
fields
Analysis of PDEs, Probability
pdf ↗ page arXiv:2511.09035 ↗

We study the generalized Hausdorff measure of some natural subsets of $k^{-1}(3)$, where $k^{-1}(3)$ consists of the real numbers $x$ for which $\left| x-\frac{p}{q} \right|<\frac{1}{(3+\varepsilon)q^2}$ has infinitely many rational solutions $\frac{p}{q}$ for any $\varepsilon<0$ but only finitely many for any $\varepsilon>0$. It is well known that $k^{-1}(3)$ is an uncountable set with Hausdorff dimension zero. Given any dimension function $h$, we determine the exact ``cut point" at which the generalized Hausdorff measure $\mathcal{H}^h(k^{-1}(3))$ drops from infinity to zero. In particular we show that such a measure is always zero or not $σ$--finite, and, as an application, we can classify topologically $k^{-1}(3)$. Moreover, we show that the subset of attainable elements of $k^{-1}(3)$ has the same generalized Hausdorff measure as $k^{-1}(3)$, but the subset of non–attainable elements of $k^{-1}(3)$ has a ``strictly smaller" generalized Hausdorff measure.

comments
Final version, revised according to the referee reports
fields
Number Theory, Dynamical Systems
pdf ↗ page arXiv:2511.10328 ↗

We analyse fully nonlinear second-order mean field games (MFG) with nondifferentiable Hamiltonians, which take the form of a coupled system of a fully nonlinear Hamilton–Jacobi–Bellman equation and a Kolmogorov–Fokker–Planck partial differential inclusion (PDI) featuring the set-valued subdifferential of the Hamiltonian. We show the existence of solutions of some stationary MFG systems with quite general coupling operators and nonnegative distributional source terms, on general bounded convex domains, under the primary assumptions of uniform ellipticity and the Cordes condition on the diffusion coefficient. The existence proof is founded on an original, and equivalent, reformulation of the PDI as a nonstandard variational inequality (VI), that offers significant flexibility in passages to limits. Furthermore, the uniqueness of the solution of the PDI/VI system is proved in the case of strictly monotone couplings. We then show how the MFG PDI/VI system in the fully nonlinear setting can be obtained as the limit of a sequence of PDE systems with differentiable Hamiltonians, and we give further results on the continuous dependence of the solution.

fields
Analysis of PDEs, Optimization and Control
pdf ↗ page arXiv:2511.12780 ↗

Reiher, Rödl, Sales, and Schacht initiated the study of relative Turán densities of ordered graphs and showed that it is more subtle and interesting than the unordered case. For an ordered graph $F$, its relative Turán density, $ρ_{<}(F)$, is the greatest $α$ such that every ordered graph $G$ has an $F$-free subgraph with at least $αe(G)$ edges. We show that $ρ_{<}(F)$ is determined by a minimal palette construction for $F$ and give a finite algorithm for computing it exactly. This has many applications: in particular we resolve two problems of Reiher, Rödl, Sales and Schacht and three problems of King et al. Moreover, in the course of the proof we show that $ρ_{<}(F)$ is equal to another natural parameter, namely the uniform Turán density of $F$, which is the infimum $α$ such that every ordered graph with density at least $α$ between any two large intervals contains a copy of $F$.

comments
28 pages
fields
Combinatorics
pdf ↗ page arXiv:2511.21571 ↗

Thin sheets, from graphene to biological membranes, exhibit wrinkling, folding, and thermal deformations driven by geometric nonlinearities. They are treated using weakly nonlinear theories such as the Föppl-von Kármán equations, whose reliance on small slopes and a single-valued height field excludes large-slope and non-graph configurations even at small strain. We introduce a geometric reformulation in terms of scalar stress and curvature potentials that preserves the FvK structure while removing the graph restriction. We show that intrinsic symmetries of these potentials, as opposed to symmetries of the configuration, can persist when none is apparent in the embedding, enabling analytical solutions of boundary-value problems inaccessible to the conventional height-based formulation. Finite-element comparisons validate the predictions under strong confinement. Together, the extended validity and tractability open analytical routes to fluctuating membranes, ribbons, and architected sheets.

journal
Physical Review Letters, 137:188201, 2026
fields
Mathematical Physics
pdf ↗ page arXiv:2512.03270 ↗

We provide a geometric condition ensuring that a very general element of a complete linear system on an abelian variety is Kobayashi hyperbolic. This can be seen as the counterpart of some classical results on very general hypersurfaces in projective spaces. Some related conjectures are also given, and involve a natural constant, the so-called basepoint-freeness threshold, associated with any polarized abelian variety $(A, L)$ detecting information about the hyperbolicity of $|L|$.

comments
16 pages, v3: final version, to appear in Bull. Lond. Math. Soc
fields
Algebraic Geometry, Complex Variables
pdf ↗ page arXiv:2512.10082 ↗

The Kittel–Shore Hamiltonian characterizes $N$ spins with identical long-range interactions, and the $\mathfrak{su}(2)$ coalgebra has been proven to be a symmetry of this model, which can be exactly solved. By using quantum groups and, in particular, $\mathfrak{su}_{q}(2)$, this Hamiltonian was deformed. In this work, we study the thermodynamic properties of this deformed model for spin-$1/2$ particles. In particular, we discuss how this deformation affects the specific heat, magnetic susceptibility, magnetisation, and phase transitions as a function of the parameter $q$ of the deformation and compare them with those of the undeformed model. Deformation was found to shift the thermodynamic behaviours to higher temperatures and alter the phase transitions. Additionally, we extend the framework to complex deformations on the unit circle, $q = e^{iθ}$, for $q$ not being a root of unity. For an $N=4$ tetrahedral cluster, this phase establishes a thermodynamic equivalence between a second-neighbor spin-Hamiltonian and the $q$-deformed quantum model. At least, we have found experimental compounds well reproduced by the Kittel–Shore model for $N=4$. Validating our effective algebraic model against exact diagonalisation data for the $\text{Cu}_2\text{Te}_2\text{O}_5X_2$ ($X=\text{Cl}, \text{Br}$) compound family demonstrates that $θ$ accurately reproduces the structural anisotropy.

comments
47 pages, 39 figures
fields
Mathematical Physics
pdf ↗ page arXiv:2512.15216 ↗

Neural networks often exhibit spectral bias, a phenomenon where high-frequency components may converge substantially more slowly than low-frequency ones. To address this imbalance, we propose a cross-attention-based network that combines a multiscale random Fourier feature bank with learnable amplitude weighting. The amplitude envelope modulates the Fourier features according to their frequencies, while the cross-attention residual structure provides an input-dependent mechanism to emphasize informative scales. As a result, the proposed design accelerates high-frequency convergence relative to comparison networks built on the same multiscale bank. Moreover, the feature bank can be incrementally enriched with dominant modes identified through discrete Fourier analysis of intermediate approximations, without modifying the hidden-layer architecture. We further extend the framework to PDE learning through a two-network representation in which one subnetwork approximates higher-frequency solution components and the other approximates lower-frequency components, with a learnable or analytically computed mixing coefficient balancing their contributions. Numerical experiments on function approximation, image reconstruction, and representative elliptic PDEs demonstrate the accuracy of the proposed method for problems involving high-frequency, multiscale, and discontinuous features.

comments
30 pages, 21 figures
fields
Numerical Analysis
pdf ↗ page arXiv:2512.18586 ↗

We prove an existence result for Robin boundary value problems modeled on \[ \begin{cases} Δu + |\nabla u|^2 + λf(x) = 0 & \text{in } Ω \\ \frac{\partial u}{\partial ν} + βu = 0 & \text{on } \partialΩ\end{cases} \] where $Ω$ is a bounded, sufficiently smooth open set in $\mathbb R^N$, $f(x)$ belongs to the Marcinkiewicz space $M^{\frac N2}$ and {$β>0$}, under a smallness assumption on the datum $λ$. In order to study such problem, we will show several properties of the weighted, singular Robin eigenvalue problem \[ λ_{1,f,γ}(Ω)= \inf_{ψ\in H^{1},\;\int_Ωfψ^{2}=1}\left\{\int_Ω|\nabla ψ|^{2}dx+γ\int_{\partialΩ}ψ^{2}\right\}. \]

comments
some proofs have been expanded, and new remarks and examples have been added
fields
Analysis of PDEs
pdf ↗ page arXiv:2512.20192 ↗

Contextual bandits are a central framework for sequential decision-making, with applications ranging from recommendation systems to clinical trials. While nonparametric methods can flexibly model complex reward structures, they suffer from the curse of dimensionality. We address this challenge using a single-index model, which projects high-dimensional covariates onto a one-dimensional subspace while preserving nonparametric flexibility. We first develop a nonasymptotic theory for offline single-index regression for each arm, combining maximum rank correlation for index estimation with local polynomial regression. Building on this foundation, we propose a single-index bandit algorithm and establish its convergence rate. We further derive a matching lower bound, showing that the algorithm achieves minimax-optimal regret independent of the ambient dimension $d$, thereby overcoming the curse of dimensionality. We also establish an impossibility result for adaptation: without additional assumptions, no policy can adapt to unknown smoothness levels. Under a standard self-similarity condition, however, we construct a policy that remains minimax-optimal while automatically adapting to the unknown smoothness. Finally, as the dimension $d$ increases, our algorithm continues to achieve minimax-optimal regret, revealing a phase transition that characterizes the fundamental limits of single-index bandit learning.

fields
Statistics Theory
pdf ↗ page arXiv:2512.24669 ↗

We introduce Heraclitean Dialectical Concept Spaces (HDCS), a topological framework for modelling stagewise structural change in systems of concepts. A stage consists of a nonempty concept set, a feasible family generating an external topology, and a neighbourhood assignment carrying additional relational structure. Neighbourhood overlap induces adjacency and channel ideals. Candidate emergence is represented by the CCER construction: an interaction profile, finite admissible cores, and anchored witnesses determine nonempty open candidate regions in uncovered remainder structure. Selected candidates may be reified as new atomic concepts at a later stage, while partial injective carry maps separately track persistence of existing identities. The resulting history space is obtained as a quotient of the disjoint union of stage spaces with the corresponding final topology. Product concept spaces extend the construction to cross-domain interaction. We derive support characterizations, mediated-interaction components, canonical CCER envelopes, monotonicity and impossibility criteria, product-space factorization, and invariance under concept-space isomorphism. Applications to monetary transformation, zero and representation, mammalian middle-ear evolution, and AI-related labour change illustrate the framework. HDCS is intended as a structural formalism rather than a causal or predictive theory of historical change.

comments
124 pages, 1 figure. Major revision of the same HDCS work. The manuscript has been extensively rewritten while retaining the same framework and research question. Revisions strengthen the foundations, reformulate CCER, clarify reification, carry maps and history space, revise applications, and add new derived structural results
fields
General Topology
pdf ↗ page arXiv:2601.00878 ↗

Stiff dynamical systems represent a central challenge in multi scale modeling across combustion, chemical kinetics, and nonlinear dynamical systems. Neural operator learning has recently emerged as a promising approach to approximate dynamical generators from data, yet stiffness imposes severe obstacles: training errors concentrate on slow manifold states, collapse of fast dynamics occurs, and the learned operator may fail to reproduce the true eigenstructure. We demonstrate three key advances enabling accurate learning of stiff operators and preserving spectral fidelity: (i) stiffness aware scaling of time derivatives, (ii) fast direction excitation via local trajectory cloud bursts, and (iii) autograd-based Jacobian diagnostics ensuring eigenstructure fidelity. Applied to the Davis-Skodje system, the approach recovers both slow and fast modes across stiffness regimes, reducing fast eigenvalue error by an order of magnitude while improving rollout fidelity. These results argue that spectral fidelity - not trajectory accuracy alone - should be a first-class target in data driven learning of stiff operators.

comments
arXiv admin note: This submission has been withdrawn because it does not meet arXiv's research content quality standards
fields
Mathematical Physics
pdf ↗ page arXiv:2601.01632 ↗

We settle an open problem regarding palindromes; that is, positive integers which are the same when written forwards and backwards. In particular, we prove that for any fixed base $b\geq 2$, there exist infinitely many square-free palindromes in base $b$. We also provide an asymptotic expression for the number of such integers $\leq x$. The core of our proof utilises a hybrid $p$-adic/Archimedean van der Corput process, used in conjunction with an equidistribution estimate of Tuxanidy and Panario, as well as an elementary argument of Cilleruelo, Luca and Shparlinski.

comments
24 pages, 1 figure, to appear in J. Lond. Math. Soc
fields
Number Theory
pdf ↗ page arXiv:2601.07097 ↗

The concepts of hypocoercivity and hypocontractivity and their relationship are studied for semi-dissipative continuous-time and discrete-time evolution equations in a Hilbert space setting. New proofs for the characterization of the short-time decay of the solution from the initial value are presented, that in particular characterize the constants in the leading terms of the solution when expanded in time. It is shown that the hypocoercivity index is equal to the hypocontractivity index when the implicit midpoint rule (the Cayley transformation) is used for the discretization. Maximally coercive/contractive representations of hypocoercive and hypocontractive semi-dissipative systems are presented, as well as the effect of different representations on the error estimates for the numerical solution.

comments
42 pages, 3 figures
fields
Dynamical Systems
pdf ↗ page arXiv:2601.09656 ↗

We present a novel matrix product representation of the Laughlin and related fractional quantum Hall wavefunctions based on a rigorous version of the correlators of a chiral quantum field theory. This representation enables the quantitative control of the coefficients of the Laughlin wavefunction times an arbitrary monomial symmetric polynomial when expanded in a Slater determinant or permanent basis. It renders the properties, such as factorization and the renewal structure, inherent in such fractional quantum Hall wavefunctions transparent. We prove bounds on the correlators of the chiral quantum field theory and utilize this representation to demonstrate the exponential decay of connected correlations and a gap in the entanglement spectrum on a thin cylinder.

comments
52 pages. Comments are welcome! Typos corrected
fields
Mathematical Physics
pdf ↗ page arXiv:2601.12165 ↗

This paper investigates model reduction methods for efficiently approximating the solution of parameter-dependent PDEs with a multi-parameter vector $\vecμ \in \mathbb{R}^p$. In cases where the Kolmogorov $N$-width decays fast enough, it is effective to approximate the solution as a sum of $N$ separable terms, each being the product of a parameter-dependent coefficient and a space-dependent function. This leads to reduced-order models with $N$ degrees of freedom and complexity of order ${\mathcal O}(N^3)$. However, when the $N$-width decays slowly, $N$ must be large to achieve acceptable accuracy, making cubic complexity prohibitive. The linear complexity measure in terms of Kolmogorov width must be replaced by the Gelfand width, with its associated sensing number. Recent nonlinear approaches based on this notion decompose the $N$ coordinates into two groups: $n$ free variables and $\overline{n}$ dependent variables, where the latter are nonlinear functions of the former ($N= n+\overline n$). Several works have focused on cases where these $\overline{n}$ functions are homogeneous quadratic forms of the $n$ variables, with optimization strategies for choosing $n$ given a target accuracy. A rigorous analysis of the local sensing number is carried out, showing that $n = p$ is optimal and appropriate, at least locally, around a reference point. In practical scenarios involving wide parameter ranges, the condition $p\le n \le p + k$ (with $k$ small) is valid and more robust from continuity arguments. Additionally, the assumption of a quadratic mapping, while justified in a local sense, becomes insufficient. More expressive nonlinear mappings-including those using machine learning-become necessary. This work contributes a theoretical foundation for such strategies and highlights the need for further investigations to push back the Kolmogorov Barrier.

fields
Numerical Analysis
pdf ↗ page arXiv:2601.13712 ↗

The monograph offers a coherent and self-contained treatment of massless (ladder) representations of the conformal group U(2,2) and their restriction to the de Sitter group Sp(2,2), combining rigorous representation-theoretic analysis with fully explicit constructions. It systematically develops these representations, including the derivation of invariant bilinear forms and Casimir operators, and constructs vertex operators and two-point functions for low-helicity fields. A central and distinctive contribution is the introduction of a canonical Clifford-split-octonion framework, in which 8-component Majorana spinors are realized within an alternative composition algebra, providing a unified and intrinsically defined setting for the algebraic, spinorial, and geometric structures underlying the theory. By bridging abstract symmetry principles with concrete computational methods and physically motivated applications in quantum field theory and cosmology, the monograph advances both conceptual clarity and technical control. While primarily addressed to researchers in mathematical physics and related fields, the exposition is carefully structured to guide advanced graduate students through subtle constructions, maintaining accessibility without compromising mathematical precision.

comments
184 pages, 6 figures. Forthcoming in the Cambridge Monographs on Mathematical Physics series, Cambridge University Press
fields
Mathematical Physics, Group Theory, Representation Theory
pdf ↗ page arXiv:2601.18433 ↗

Mixed-precision computations are a hallmark of the current stage of AI, driving the progress in large language models towards efficient, locally deployable solutions. This article addresses the floating-point computation of compositionally-rich functions, concentrating on transformer inference. Based on the rounding error analysis of a composition $f(g(\mathrm{x}))$, we provide an adaptive strategy that selects a small subset of components of $g(\mathrm{x})$ to be computed more accurately while all other computations can be carried out with lower accuracy. We then explain how this strategy can be applied to different compositions within a transformer and illustrate its overall effect on transformer inference. We study the effectiveness of this algorithm numerically on GPT-2 models and demonstrate that already very low recomputation rates allow for improvements of up to two orders of magnitude in accuracy.

comments
Accepted in NeurIPS 2026
fields
Numerical Analysis
pdf ↗ page arXiv:2601.21623 ↗

In this paper, we investigate the regularization effect of fractional stochastic forcing on Burgers-type equations with fractional dissipation, with an application to the Degasperis–Procesi (DP) equation. In particular, we consider the perturbation induced by the singular noise $|D|^{1/2}ξ$ and establish local well-posedness in the negative Sobolev space $H^{-1/4+δ}$ for some small $δ>0$. Due to the singular nature of the nonlinear interactions, classical solution theories cannot be directly applied. Inspired by the framework developed in \cite{hairer2013solving,gubinelli2017kpz}, we introduce a generalized solution theory based on an enhanced structure and derive the effective equations satisfied by these generalized solutions. Our main contribution is the establishment of a general framework for describing singular PDEs driven by rough data. Moreover, we prove the convergence of the associated non-Gaussian rough structures in the fractional dissipation setting. As an application, we apply this framework to the stochastic Degasperis–Procesi equation and obtain its local well-posedness in the low-regularity regime.

fields
Analysis of PDEs, Probability
pdf ↗ page arXiv:2602.07492 ↗

We establish uniform a priori estimates for solutions of semilinear planar Hamiltonian elliptic systems in a ball with Dirichlet boundary conditions. We consider a broad class of coupled nonlinearities with asymptotic critical behaviour in the sense of Brezis–Merle. The approach we follow is based on a blow-up analysis combined with Liouville–type theorems and integral estimates. Our results extend the scalar theory of uniform a priori bounds to the Hamiltonian case, and solve an open problem in [de Figueiredo D.G., do Ó J.M., Ruf B., Adv. Nonlinear Stud. 6 (2006), no. 2]. We believe that this approach is new in this setting. As a consequence of our a priori estimates, we prove the existence of a positive solution by means of fixed-point index theory.

comments
The argument of the proof in Section 4.2 has been adjusted; the main results proved in Section 3 have been collected in a unified proposition; assumption (H_2) is now global; the proofs of the preliminarylemmas have been postponed in a dedicated Appendix; several minor typos and style issues have been fixed
fields
Analysis of PDEs
pdf ↗ page arXiv:2602.11720 ↗

Some non-trivial upper bounds are proved relating to the triple product property (TPP) for subgroups of finite nilpotent groups of class $2$ using elementary group theory. The TPP is the property defined for three non-empty subsets $S, T, U$ of a group $G$ that the group equation $s's^{-1}t't^{-1}u'u^{-1} = 1$, over pairs of elements $s', s \in S$, $t', t \in T$, $u', u \in U$, is satisfied if and only if $s' = s$, $t' = t$, $u' = u$. When $G$ is finite, and the parameter $ρ_0(G)$, called \emph{subgroup TPP ratio}, is defined as $ρ_0(G) := \max\frac{|S||T||U|}{|G|}$, where the maximum is over the collection of all triples of subgroups $S, T, U$ of $G$ satisfying the TPP, this paper proves that \textup{(1)} $ρ_0(G) < \sqrt{G:Z(G)}$} for (all) groups of nilpotency class $2$, \textup{(2)} $ρ_0(G) \leq p$ for $p$-groups with a cyclic commutator subgroup of order $p$, \textup{(3)} $ρ_0(G) = 1$ for $p$-groups of nilpotency class $2$ with a "large" centre such that either $G/Z(G)$ is isomorphic to $C_p \times C_p$ or $C_{p^2} \times C_p$, \textup{(4)} and $ρ_0(G) = 1$ for $p$-groups of nilpotency class $2$ with "small" (irreducible, complex) character degrees of $1$ or $p$ where $G/Z(G)$ is not isomorphic to $C_p \times C_p \times C_p$.

comments
Remove extraneous MSC code, plus minor textual and typographical improvements
fields
Group Theory
pdf ↗ page arXiv:2602.15796 ↗

This paper is devoted to the study of the falsification by fellow-traveller property (FFTP) in Dyer groups. We exhibit a finite generating set for which the associated Cayley graph is a locally finite mediangle graph, and leverage its properties to prove that Dyer groups have the FFTP with respect to this generating set, from which it follows that Dyer groups have finitely many cone types. A key component of the proof is a generalisation of the Parallel Wall Theorem from Coxeter groups to Dyer groups. These results emphasise the role of Dyer groups in providing a unified approach to Coxeter groups and graph products of cyclic groups.

comments
33 pages, 2 figures. Comments are welcome!
fields
Group Theory
pdf ↗ page arXiv:2602.18220 ↗

It is well known that the order of the torsion group of an abelian variety over a number field divides the greatest common divisor of the sizes of points on the reduction at each prime (except for an explicit finite set). Drawing inspiration from Katz's Inventiones paper (1981), we investigate the converse to this for abelian varieties of \(\rm GL_2\)-type and exhibit a tentative list of possible rational torsion orders for modular abelian varieties over \(\mathbb Q\) of dimension up to \(5\).

comments
12 pages, comments are welcome! Updated version for resubmission. Github repo: https://github.com/NirvanaC93/Torsion-of-GL2-abelian-varieties
fields
Number Theory
pdf ↗ page arXiv:2602.21047 ↗

By exploiting the arithmetic homotopy of the moduli spaces of curves, Galois-Teichmüller theory stands at the interface of braid-mapping class groups and of anabelian geometry. Starting from the classical braid-theoretic construction of the Grothendieck-Teichmüller group, we review how anabelian geometry -- beginning with the foundational work of Nakamura -- provides the arithmetic mechanisms underlying its definition. We then explain how the combinatorial anabelian geometry developed by Hoshi and Mochizuki recasts these constructions within a purely group-theoretic and algorithmic framework. In particular, we describe how the group GT emerges as an anabelian object and how, once freed from an auxiliary or artificially imposed containers, the anabelian algorithms yield a combinatorial reconstruction of the absolute Galois group of rational numbers. The perspective developed here highlights a conceptual shift from explicit braid-theoretic computations to functorial and algorithmic forms of anabelian reconstruction.

comments
20 pages, 17 figures; New: precision and comments; Accepted for publication in: Low Dimensional Topology, Number Theory and Arithmetic Galois Theory: In honor of Hiroaki Nakamura's 60th birthday
fields
Algebraic Geometry
pdf ↗ page arXiv:2603.02848 ↗

We consider convex mean-field vector spin glasses perturbed by a general Mattis interaction $N G(m_N)$, where $m_N$ is a bounded vector-valued empirical average and $G$ is continuous. We prove a Parisi-type variational formula for the limiting quenched free energy and a quenched large deviation principle for $m_N$, both at every temperature. The approach is simple and modular: the only model-specific spin-glass input is the variational formula for every linear tilt of $m_N$ together with differentiability of the resulting free energy. The classical Gärtner–Ellis, Varadhan, and Bryc transfers then treat an arbitrary continuous $G$. Thus the same argument applies to a broad class of empirical observables and Mattis potentials, without a separate constrained-free-energy calculation for each model. As a scalar specialization, the framework contains the Gaussian spin-glass surrogate used in mismatched rank-one matrix estimation.

comments
15 pages
fields
Probability
pdf ↗ page arXiv:2603.12033 ↗

We study enriched mean-field vector spin glasses whose covariance function need not be convex. In a quantitative short-time regime, corresponding to sufficiently high temperature in the usual physical parameterization, we prove convergence of the quenched free energy to the differentiable solution of a infinite-dimensional Hamilton–Jacobi equation proposed for non-convex models. We identify that solution through its characteristics and a unique fixed point. The argument does not assume replica symmetry. We then allow a general continuous Mattis interaction $N G(m_N)$, obtain the corresponding variational formula, and prove a quenched large deviation principle for $m_N$. As an explicit application, we treat a bipartite model with Mattis fields, which is closely related to the Restricted Boltzmann Machine. The application requires neither a replica-symmetry assumption nor ellipticity or Nishimori identities; its restriction is the same quantitative high-temperature condition as in the general theorem.

comments
39 pages
fields
Probability
pdf ↗ page arXiv:2603.12034 ↗

We give a proof of the o-minimal version of the Whitney Extension Theorem simplified as compared to the original ones. A new simplifying ingredient is a definable variant of Urysohn's lemma for class $\mathcal{C}^q$ (see Section 3).

comments
Minor corrections, to be published in Studia Mathematica
fields
Logic, Algebraic Geometry
pdf ↗ page arXiv:2603.14924 ↗

We take quantum theory and replace C by C[e] where e^2 = 0, i.e. we extend quantum theory to the ring of dual complex numbers. The aim is to develop a common language in which to treat continuous quantum physics and discrete quantum models in a unified manner, including their symmetries. Since quantum theory is linear, introducing e is enough to model infinitesimals. A first objection to this programme is that C[e] is not a field, since division by e is undefined, while quantum mechanics typically relies on division. A second objection concerns whether unitarity still makes sense given e^2 = 0. Hence, the core of this work is dedicated to proving that dual-complex quantum theory remains fully consistent. In particular, norm is preserved at all times, and renormalization never requires dividing by an infinitesimal. An equivalence with conventional quantum theory is demonstrated: the dual-complex extension of a parametrized quantum operation automatically provides a linear treatment of its first-order variations. As a first example application, we provide a unified description of both the Dirac equation in the continuum and the Dirac Quantum Walk in the discrete. We establish the discrete Lorentz covariance of the latter.

comments
In Proceedings QPL 2026, arXiv:2611.04069
journal
EPTCS 456, 2026, pp. 166-187
fields
Quantum Algebra
pdf ↗ page arXiv:2603.17581 ↗

Given a pivotal module category over a spherical fusion category, we introduce the encircling module, a module over the fusion algebra defined using the pivotal structure, and prove that it is isomorphic to the NIM-rep as a fusion algebra module. When applied to the $\mathcal{TM}$ realisation of the modular invariant partition function (arXiv:1911.09024), this identifies the NIM-rep multiplicities with the entries $Z_{I,I^\vee}$ of the modular invariant (its diagonal relative to the charge-conjugation invariant, which is the one given by the regular module category), providing a categorical generalisation of Böckenhauer, Evans and Kawahigashi's results (arXiv:math/9907149). We also show that for indecomposable module categories the dimension condition on $\mathcal{TM}$ required for modular invariance is automatically satisfied, and that $\mathcal{TM}$ is, as an algebra, the full centre of the module category in the sense of Davydov, Kong and Runkel (arXiv:1307.5956), which for module categories of algebras is the full centre of Fjelstad, Fuchs, Runkel and Schweigert (arXiv:hep-th/0612306). Consequently $\mathcal{TM}$ gives rise to Cardy algebras in the sense of Kong and Runkel (arXiv:0807.3356).

comments
23 pages, many figures
fields
Quantum Algebra, Mathematical Physics
pdf ↗ page arXiv:2603.20545 ↗

We prove determinantal-Pfaffian formulae that simultaneously generalise the Pfaffian minor summation formula of Ishikawa and Wakayama and Byun's recent minor summation formula. These formulae are based on factorisation formulae for the determinant of the sum of a skew-symmetric matrix and a rank-1 matrix. Applications include a Cauchy-type identity for skew Schur functions.

comments
17 pages, AmS-LaTeX; corrected the sign calculations in Propositions 7-9, and made further, minor corrections
fields
Combinatorics
pdf ↗ page arXiv:2603.21021 ↗

We give a practical construction of functions of $N$ variables that are simultaneously equivariant under a group action and invariant under permutations, or simply group-equivariant, valid for any finite-dimensional space stable under that action. The method applies to any connected linear Lie group: it uses the Lie algebra to build a matrix $M$ whose kernel corresponds exactly to the subspace with the desired symmetry properties, removing the need for Clebsch-Gordan coefficients and explicit symmetrization over permutations. For the groups $SO(3)$ and $SU(2)$, the special structure of the matrix $M$ yields a linear scaling method which significantly outperforms the typically exponential cost of existing methods in the literature. As a byproduct, we obtain the exact dimension of group-equivariant and permutation-invariant spaces. This allows us to compare dimensionalities and show that, for large $N$, this dimension is of comparable order to that of the permutation-invariant space alone, while pre-asymptotically it is orders of magnitude lower. Enforcing group-equivariance explicitly, on top of permutation-invariance, therefore yields a substantial computational gain when approximating such functions.

comments
42 pages, 9 figures, 5 tables
fields
Numerical Analysis
pdf ↗ page arXiv:2604.01975 ↗

We study a catching-up algorithm with approximate projections for finite-dimensional differential inclusions governed by a maximally monotone operator and a continuous perturbation. Through the canonical decomposition of the operator into a regular part and the normal cone to the closure of its domain, the problem becomes a sweeping process, discretized by an explicit predictor step and a feasible approximate projection. For variable time steps, the approximations converge, up to a subsequence, to a solution on every finite horizon when the mesh size and the accumulated projection-error budget vanish and the nodes stay in a common compact convex set. This set comes from linear growth on finite horizons and, without any growth condition, by dissipativity of the projected field at a feasible center, which confines the nodes to a fixed ball uniformly in time for small steps and small projection errors. Under either condition, locally Lipschitz perturbations give global well-posedness, convergence of the whole family, and, for a common initial state, a trajectory-error estimate of order $O(\sqrt{h+E})$, where $h$ is the maximal step and $E$ the accumulated budget; sparse projection errors are allowed. An elastic network with unilateral stops illustrates the results, and numerical experiments compare the algorithm with the full-resolvent discretization.

fields
Optimization and Control
pdf ↗ page arXiv:2604.10774 ↗

Inductive bias refers to restrictions on the hypothesis class that enable a learning method to generalize effectively from limited data. A canonical example in control is linearity, which underpins low sample-complexity guarantees for stabilization and optimal control. For general nonlinear dynamics, by contrast, guarantees often rely on smoothness assumptions (e.g., Lipschitz continuity) which, when combined with covering arguments, can lead to data requirements that grow exponentially with the ambient dimension. In this paper we argue that data-efficient nonlinear control demands exploiting inductive bias embedded in nature itself, namely, structure imposed by physical laws. Focusing on Hamiltonian systems, we leverage symplectic geometry and intrinsic recurrence on energy level sets to solve target reachability problems. Our approach combines the recurrence property with a recently proposed class of policies, called chain policies, which composes locally certified trajectory segments extracted from demonstrations to achieve target reachability. We provide sufficient conditions for reachability under this construction and show that the resulting data requirements depend on explicit geometric and recurrence properties of the Hamiltonian rather than the state dimension.

comments
12 pages
fields
Optimization and Control
pdf ↗ page arXiv:2604.17213 ↗

We present the Multi-Block DC (BDC) class, a rich class of structured nonconvex functions that admit a DC ("difference-of-convex") decomposition across parameter blocks. This multi-block class not only subsumes the usual DC programming, but also turns out to be provably more powerful. Specifically, we demonstrate how standard models (e.g., polynomials and tensor factorization) must have DC decompositions of exponential size, while their BDC formulation is polynomial. This separation in complexity also underscores another key aspect: unlike DC formulations, obtaining BDC formulations for problems is vastly easier and constructive. We illustrate this aspect by presenting explicit BDC formulations for modern tasks such as deep ReLU networks, a result with no known equivalent in the DC class. Moreover, we complement the theory by developing algorithms with non-asymptotic convergence theory, including both batch and stochastic settings, and demonstrate the broad applicability of our method through several applications.

comments
35 pages, 4 Figures
fields
Optimization and Control
pdf ↗ page arXiv:2604.17560 ↗

We prove the semistable reduction theorem for $\mathcal{E}^†_K$-valued and $K$-valued overconvergent $F$-isocrystals over $k((t))$-varieties which were introduced by Lazda and Pál. As an application, we prove the finite dimensionality of $\mathcal{E}^†_K$-valued rigid cohomology with compact support.

fields
Number Theory, Algebraic Geometry
pdf ↗ page arXiv:2604.17799 ↗

We prove a dimension-explicit upper bound for the information-complexity (also called the inverse of the discrepancy) of the classical spherical cap $L_2$ discrepancy on the $d$-dimensional sphere $\mathbb{S}^d$. For every fixed $\varepsilon\in(0,1)$, this bound implies that a single point suffices in all sufficiently large dimensions, which shows a ``blessing of dimensionality'' for the associated numerical integration problem. We then introduce a one-parameter family of distance reproducing kernels for the Sobolev space $H^{(d+1)/2}(\mathbb{S}^d)$. For the balanced kernel $K(\boldsymbol{x},\boldsymbol{y})=1-\tfrac{1}{\sqrt{2}}\|\boldsymbol{x}-\boldsymbol{y}\|$, a lower bound together with a random-point upper bound shows that the normalized information-complexity is $Θ_\varepsilon(d)$ for every fixed $\varepsilon\in(0,1)$. Thus its dependence on $d$ is determined up to constants depending only on $\varepsilon$. Finally, we study the one-sided spherical cap discrepancy obtained by restricting the height parameter to $[0,1]$. We derive its kernel and Stolarsky's invariance principle explicitly. A sharp mean-distance inequality, attained by regular simplices, determines its normalized information-complexity exactly in all sufficiently large dimensions and proves strong polynomial tractability with optimal $\varepsilon$-exponent two.

fields
Numerical Analysis, Number Theory
pdf ↗ page arXiv:2604.21340 ↗

We study the generalized Lam'e equation (GLE) on an elliptic curve $E$ with multiple regular singularities $\mathbf{p} = (p_i)_{i = 1}^r$ of weights $\mathbf{n} = (n_i)_{i = 1}^r$. By analyzing the locus admitting quasi-periodic solutions, we construct two fundamental algebraic curves: (i) The generalized Lam'e curve (GLC), $\mathcal{Y}_{\mathbf{n}, \mathbf{p}}$, which lies in an affine bundle over $\operatorname{Sym}^n E$ for total weight $n:=\sum n_i \in \mathbb{Z}_{\geq 0}$ and parametrizes generalized Hermite–Halphen ansatz solutions. (ii) The log-free curve, $V_{\mathbf{n}, \mathbf{p}}$, a non-complete intersection variety arising when all $n_i \in \frac{1}{2}\mathbb{N}$, which we prove is a reduced curve, confirming a conjecture of Wang. We analyze the GLC as an algebraic family over the pole configuration space. By studying the addition map$$σ\colon \operatorname{Sym}^n E \longrightarrow E,$$where we establish a generically finite, universal degree formula, we show that the geometry of boundary degenerations under pole collisions perfectly mirrors the tensor algebra of $\mathfrak{sl}_2(\mathbb{C})$-modules within the BGG category $\mathcal{O}$. This provides the local structural limits needed to establish the global flatness of the GLC. Furthermore, we develop a framework of twisted isomonodromic deformations and construct $(\mathbf{n}, \mathbf{p})$-deformed pre-modular forms parameterized by twisted monodromy data $(t,s)$. Their vanishing solves the underlying monodromy problem and factorizes along boundary strata, allowing an arbitrary configuration to be continuously deformed down to the classical Lam'e equation. Finally, using an asymptotic scaling technique, we give a complete extension of Treibich's generic counting for $r=2$ symmetric pairs to all smooth elliptic curves.

comments
84 pages; revised final section and minor corrections throughout; comments welcome
fields
Algebraic Geometry, Classical Analysis and ODEs
pdf ↗ page arXiv:2604.21880 ↗

We investigate the large scale geometry of certain metric spaces through the lens of dynamics. Our approach establishes a close connection between large scale dynamical phenomena and operator algebras by characterizing various large scale dynamic behaviors in terms of GNS representations of the uniform Roe algebras arising from natural canonical states. Our dynamical systems are given by the Stone-Čech boundary of metric spaces together with their inverse semigroup of partial translations. This defines a space of orbits and we characterize Hausdorffness and $T_1$-ness of this space by the failure of coarse embeddability of certain metric spaces. Surprisingly, while the orbit space has very weak separation properties, we show that it satisfies a certain ''localized version'' of Urysohn's lemma. We show that the topology of the space of orbits and quasi-orbits are given by the space of irreducible representations of uniform Roe algebras and by the space of their primitive ideals, respectively. As a highlight of the theory developed herein, we provide classes of spaces such that the prime ideals of their uniform Roe algebras are primitive. This is the case for instance of spaces whose orbit space is $T_1$.

comments
55 pages
fields
Operator Algebras, Dynamical Systems, Functional Analysis
pdf ↗ page arXiv:2604.24452 ↗

We construct a complex of toric varieties we call the quasisymmetric Grassmannian inside the Grassmannian of $r$-planes in $\mathbb{C}^n$. Each irreducible component is a positroid variety and an $S_n$ translate of a toric Richardson variety of ribbon shape. We describe it as the vanishing locus of equations $Δ_AΔ_{A'}=0$ in Plücker coordinates determined by a new noncrossing combinatorial object we call the quasisymmetric Johnson graph. We give an affine paving, and show that its cohomology ring is a quasisymmetric modification of the Borel presentation of the Grassmannian's cohomology, with fundamental quasisymmetric polynomials playing the role of Schur polynomials.

comments
35 pages, fixing small typos
fields
Algebraic Geometry, Combinatorics
pdf ↗ page arXiv:2604.24903 ↗

We construct a measure in the hamiltonian function level sets that is invariant under the hamiltonian flow for short times and flow preserving for arbitrarily long times. This allows a probabilistic approach to the study of hamiltonian systems, in the space of states with fixed energy. We prove that this measure generates the microcanonical partition function employed in physics and show that it can be transformed into the canonical partition function in an asymptotic limit, hence reproducing classical Statistical Physics. We also argue that this gives an alternative solution to Simon's second problem.

fields
Mathematical Physics
pdf ↗ page arXiv:2604.25066 ↗

We investigate finite time dynamics in nonlinear classical scalar field systems containing one ghost and one normal kinetic sector. Using a spacetime finite element formulation, we study the dependence of threshold crossing and numerical continuation times on initial amplitude, characteristic wavenumber, spectral tilt, phase structure, and spatial localization. In the quartic model, larger amplitudes generally lead to earlier growth, while higher wavenumbers and ultraviolet-biased spectra favor longer resolved bounded evolution. Smooth compact initial data exhibit a broad region with no observed growth threshold crossing over the simulated interval. A direct comparison of two Gaussian evolutions shows that enhanced low-wavenumber power can precede rapid field growth, although this observation does not establish a universal spectral-transfer mechanism. We also compare the quartic interaction with a lifted $φ^6$ potential. The latter gives a markedly different continuation landscape, with a broad long-converged region followed by shorter accessible terminal times at larger amplitudes. Localization diagnostics do not identify a persistent oscillon core in the selected cases. By distinguishing amplitude growth, spectral reorganization, solver-limited termination, and finite-time censoring, our results characterize a nontrivial metastability landscape without implying global or quantum stability.

comments
28 pages (including 14 figures and 6 tables) with 3 appendices and references
fields
Dynamical Systems
pdf ↗ page arXiv:2604.25635 ↗

We introduce Hyper Input Convex Neural Networks (HyCNNs), a novel neural network architecture designed for learning convex functions. HyCNNs combine the principles of Maxout networks with input convex neural networks (ICNNs) to create a neural network that is always convex in the input, theoretically capable of leveraging depth, and performs reliably when trained at scale compared to ICNNs. Concretely, we prove that HyCNNs require exponentially fewer parameters than ICNNs to approximate quadratic functions up to a given precision. Throughout a series of synthetic experiments, we demonstrate that HyCNNs outperform existing ICNNs and MLPs in terms of predictive performance for convex regression and interpolation tasks. We further apply HyCNNs to learn high-dimensional optimal transport maps for synthetic examples and for single-cell data, where they oftentimes outperform ICNN-based neural optimal transport methods and other baselines across a wide range of setting.

comments
71 pages, 17 figures, the first two authors contributed equally
fields
Statistics Theory
pdf ↗ page arXiv:2604.26942 ↗

We introduce a highly general method to determine the energy minimum of spin Hamiltonians over separable states when the single-particle reduced density matrices are fixed. We present efficient numerical methods that reduce the problem to calculations in a two-particle system. For a broad class of very general spin models, we present analytic formulas for the minimum. For ferromagnetic Ising and Ising-like models of spin-j particles, the minimum over separable states on two-colorable lattices, and over symmetric separable states on fully connected graphs, is given by a compact analytic formula involving the quantum Fisher information. For antiferromagnetic Ising-like and XY-like models of spin j-particles living on a complete graph, the energy minimum for symmetric separable states is given with the expectation value of single-particle operators. For the ferromagnetic Heisenberg chain of spin-1/2 particles, the minimum is expressed via the Uhlmann-Jozsa fidelity. These relations enable the extraction of the quantum Fisher information and quantum fidelity of the corresponding single-particle reduced states from correlation measurements on ground-state ensembles of suitably engineered spin models. For a number of other models, we obtain not necessarily tight lower bounds. We also present bounds for various levels of multiparticle entanglement.

comments
9 pages including 3 figures + 11-page Supplement, revtex 4.2
fields
Mathematical Physics
pdf ↗ page arXiv:2605.03022 ↗

We prove relative smoothing with uniform reach control for one-parameter families of positive-reach $C^{1,1}$ knots and combine it with relative multijet transversality to obtain projection-generic regularization inside strict ropelength sublevels. Passing to a lifted Reidemeister multigraph reconstructs the admissible $π_0$ of every strict projection-framed ropelength sublevel, and the right-relaxed diagrammatic merge scale equals the geometric one. The normalised projection-framed sublevels are compact, so connected merge levels are attained. These degree-zero results are organized by a fibre cosheaf: a filtered relational cellular cosheaf of sets on the planar Reidemeister multigraph, whose stalks are strict fibre components and whose costalks are the relations realized by Reidemeister movies. The lifted multigraph is its Grothendieck construction, reconstruction is a descent statement, and the diagram image is its support. The diagram image computes merges exactly when no stalk is persistently separated, and a canonical flat coarsening is the coarsest diagram-indexed model that computes components. For amphichiral knots the spherical diagram image cannot detect mirror barriers, whereas the planar image can; no planar type of the standard diagrams of $4_1$ and $6_3$ is fixed by crossing switch. Symmetric hubs and finite certificates give upper bounds for mirror merges, and sign separators give lower bounds. For finite recognition we distinguish full rooted-ball occurrence, visibility in the monotone diagram-image filtration, and saturation. Every finite Reidemeister submultigraph becomes visible at a finite ropelength level, a visible saturated Barbensi-Celoria characteristic certificate determines the knot type up to mirroring, and a universal computable crossing-complexity bound controls visibility.

comments
75 pages, 13 figures. v4: new Section 6 on the fibre cosheaf (lifted multigraph as Grothendieck construction, reconstruction as descent, when diagram images compute merges); new Section 10.4 on mirror merges with computer-verified results for 4_1 and 6_3 (scripts in ancillary files). Sections 6-11 of v3 are now 7-12
fields
Geometric Topology, Metric Geometry
pdf ↗ page arXiv:2605.03350 ↗

Deep neural networks exhibit periodic loss spikes during unregularized long-term training, a phenomenon known as the "Slingshot Mechanism." Existing work usually attributes this to intrinsic optimization dynamics, but its triggering mechanism remains unclear. This paper proves that this phenomenon is a result of floating-point arithmetic precision limits. As training enters a high-confidence stage, the difference between the correct-class logit and the other logits may exceed the absorption-error threshold. Then during backpropagation, the gradient of the correct class is rounded exactly to zero, while the gradients of the incorrect classes remain nonzero. This breaks the zero-sum constraint of gradients across classes and introduces a systematic drift in the parameter update of the classifier layer. We prove that this drift forms a positive feedback loop with the feature, causing the global classifier mean and the global feature mean to grow exponentially. We call this mechanism Numerical Feature Inflation (NFI). This mechanism explains the rapid norm growth before a Slingshot spike, the subsequent reappearance of gradients, and the resulting loss spike. We further show that NFI is not equivalent to an observed loss spike: in more practical tasks, partial absorption may not produce visible spikes, but it can still break the zero-sum constraint and drive rapid growth of parameter norms. Our results reinterpret Slingshot as a numerical dynamic of finite-precision training, and provide a testable explanation for abnormal parameter growth and logit divergence in late-stage training.

comments
29 pages, 13 figures; accepted to NeurIPS 2026
fields
Optimization and Control
pdf ↗ page arXiv:2605.06152 ↗

Establishing almost sure convergence rates for stochastic approximation and reinforcement learning under Markovian noise is a fundamental theoretical challenge. We make progress towards this challenge for a class of stochastic approximation algorithms whose expected updates are contractive, a setting that arises in many reinforcement learning algorithms such as $Q$-learning and linear temporal difference learning. Specifically, for a power-law learning rate $O(n^{-η})$ with $η\in (1/2, 1)$, we obtain an almost sure convergence rate arbitrarily close to $o(n^{1 - 2η})$. For the harmonic learning rate $α_n=α/(n+1)$, the almost sure squared-error rate is $o(n^{-ζ})$ for every $ζ<\min\{1,2(1-κ)α\}$, where $κ$ is the contraction factor. Thus exponents arbitrarily close to one are available when $2(1-κ)α\geq1$. This approaches the $n^{-1}$ polynomial exponent of classical law-of-the-iterated-logarithm benchmarks. Key to our analysis is a novel Lyapunov drift construction that applies a Poisson-equation based correction for Markovian noise to the well-established Moreau-envelope smoothing for the contractive mapping.

fields
Optimization and Control
pdf ↗ page arXiv:2605.07104 ↗

Inverse optimization (IO) seeks to infer the parameters of a decision-maker's objective from observed context–action data. We study noiseless IO, where demonstrations are generated by a ground-truth objective. We provide a high-probability ${O}(\frac{d}{T})$ generalization bound for the induced action set, where $d$ is the number of unknown parameters and $T$ is the size of the training dataset. We strengthen this guarantee under additional conditions that ensure uniqueness of the chosen action, bringing our IO guarantees in line with best-arm identification results in the bandit literature. We further show that the ${O}(\frac{d}{T})$ rate is tight over all consistent estimators considered here, and extend the result to both instantaneous and cumulative regret. Notably, the resulting regret lower bound matches the corresponding upper bounds in the adversarial setting, indicating that the stochastic IO setting is effectively adversarial for the class of estimators studied here. Finally, we propose a parameter-free algorithm with lower per-iteration complexity than generic solvers. Experiments validate the predicted rates and illustrate the tightness of our bounds.

comments
21 pages, 2 figures
fields
Optimization and Control
pdf ↗ page arXiv:2605.08866 ↗

We investigate conformal relative equilibria for Hamiltonian systems on exact Poisson manifolds equipped with scaling symmetries. By introducing conformally Poisson actions and conformal momentum maps, we characterize these equilibria through an augmented Hamiltonian formulation; in the nondegenerate case, this recovers the conditions recently developed for the exact symplectic case. Specializing to the Lie-Poisson manifold $\mathfrak g^*$ associated with an arbitrary finite-dimensional real Lie algebra $\mathfrak g$, we show that its natural scaling action makes $\mathfrak g^*$ an exact Poisson manifold. We then establish a purely algebraic existence criterion: there exists a homogeneous Hamiltonian on $\mathfrak g^*$ admitting a nontrivial conformal relative equilibrium if and only if $\mathfrak g$ contains an element whose adjoint operator has a nonzero real eigenvalue. Applying this criterion to the Bianchi classification, we identify exactly the real three-dimensional Lie algebras for which there exists a homogeneous Hamiltonian admitting a nontrivial conformal relative equilibrium. As a prominent application, we show that nontrivial conformal relative equilibria exist for a rigid body on $\mathfrak{so}(2,1)^*$, whereas the classical free rigid body on $\mathfrak{so}(3)^*$ admits no nontrivial conformal relative equilibria.

journal
SIGMA 22 (2026), 108, 26 pages
fields
Mathematical Physics, Differential Geometry, Dynamical Systems
pdf ↗ page arXiv:2605.09062 ↗

\emph{Kullback-Leibler} (KL) regularization is ubiquitous in reinforcement learning algorithms in the form of \emph{reverse} or \emph{forward} KL. Recent studies have demonstrated $ε^{-1}$-type fast rates for decision making under reverse KL regularization, in contrast to the standard $ε^{-2}$-type sample complexity. However, for forward-KL-regularized objectives, existing statistical analyses are either not applicable or result in $\tilde{O}(ε^{-2})$ slow rates. We take the first step towards addressing this problem via a streamlined analysis of forward-KL-regularized offline CBs. We give the first $\tilde{O}(ε^{-1})$ upper bounds in tabular and general function approximation settings, both under notions of \emph{single-policy concentrability}. In particular, our convex-analytical pipeline unifies these settings by exploiting the pessimism principle in a novel way and completely bypasses the proof routines in previous works based on the mean value theorem, which might be of independent interest. Moreover, we provide rate-optimal lower bounds, manifesting the tightness of our upper bounds in terms of statistical rates. Our lower bounds also demonstrate that the forward-KL-regularized sample complexity recovers the unregularized slow rate in the low-regularization regime, similarly to the reverse-KL regularization.

comments
36 pages, camera-ready version
fields
Information Theory, Statistics Theory
pdf ↗ page arXiv:2605.09214 ↗

Minimal filling architectures describe the smallest hidden-layer widths for which the neurovariety of a polynomial neural network fills its ambient function space. It was conjectured that every such architecture has unimodal hidden widths. We construct counterexamples to this conjecture using a frontier search and certify the main example through finite-field Jacobian computations and recursive dimension bounds. In the opposite direction, we prove that every minimal filling architecture of depth at most three is unimodal. A nonunimodal minimal filling architecture of depth four shows that the depth bound is sharp.

comments
26 pages, 3 figures, 3 tables. Added results on low depth architectures, more counterexamples, rewritten appendix, and updates to exposition
fields
Algebraic Geometry
pdf ↗ page arXiv:2605.09609 ↗

We study the $L^p$ concentration problem for the Born-Jordan distribution in dimension $d>1$, thus extending the one-dimensional analysis in [Stra-Svela-Trapasso, J. Math. Pures Appl. (2026)]. We show that the existence of concentration optimizers depends on the exponent $p$ with a critical threshold at $p_*(d)= \frac{2d}{d-2}$ for $d\geq2$ (with the understanding that $p_*(2)=\infty$). In particular, for subcritical exponents $1\leq p<p_*(d)$ we prove that the supremum is finite and is attained, whereas for supercritical exponents $p>p_*(d)$ we show that the functional is unbounded. We also provide the complete solution in the (significantly more) challenging critical regime in dimension $d=2$. For the critical case in dimension $d=2$ and the supercritical case in dimension $d>2$ we also show that $L^2$ functions whose Born-Jordan distribution does not belong to $L^p$ are generic in the Baire sense. For the critical exponent in dimension $d>2$, for which Wang has recently shown that the supremum is finite, we prove that the functional is not sequentially weakly upper semicontinuous, showing that the still open question of attainment is nontrivial.

comments
29 pages. Analysis of the critical case in high dimension has been added
fields
Classical Analysis and ODEs, Functional Analysis
pdf ↗ page arXiv:2605.10535 ↗

We construct moment functions that are Neyman-orthogonal to a chosen order in moment-condition models that involve finite-dimensional nuisance parameters. These moment functions reduce sensitivity to estimation error in the nuisance parameters and, as such, offer a unified and tractable route to higher-order debiasing in a wide range of econometric models. The number of additional nuisance parameters required by our construction, beyond those already present in the original moment conditions, is independent of the order of orthogonalization and can be reduced to a single scalar if desired.

fields
Statistics Theory
pdf ↗ page arXiv:2605.10842 ↗

We study the solitons of the inverse curve shortening flow in the hyperbolic plane with respect to Killing fields: all parabolic fields, and the hyperbolic and elliptic fields with unit translation length and unit angular speed. The three cases reduce to one planar system. Except for a family of horocycles, every soliton has an endpoint at finite distance where the curvature blows up and the curve is a branch of a cusp. In the parabolic and elliptic cases there is a critical value $H$ of the parameter of the family: below $H$ the other end meets the ideal boundary orthogonally, above $H$ it is a second singular endpoint, and at $H$ the curvature tends to $1$. In the hyperbolic case $H=π/2$ and the critical solitons are horocycles. These horocycles are the only complete solitons and the only solitons of constant curvature.

comments
Substantially revised and extended version: all three classes of Killing fields are now treated, with new results on the critical soliton and complete solitons. 20 pages, 3 figures
fields
Differential Geometry
pdf ↗ page arXiv:2605.14385 ↗

In this paper, a class of linear authentication codes with secrecy, which are equipped with simple encoding rules, is constructed. By means of a special Weil sum, the maximum success probabilities of impersonation attack (denoted by $P_I$) and of substitution attack (denoted by $P_S$) for these codes are explicitly derived. It is further proven that the codes are asymptotically optimal, in the sense that both the information theoretic bound and the combinatorial bound for authentication codes are asymptotically attained by $P_I$ and $P_S$.

fields
Information Theory
pdf ↗ page arXiv:2605.14823 ↗

We introduce the notions of $rgs$ and $irgs$ for Keisler measures, motivated by the study of generically stable random types and their associated Morley sequences. We obtain characterizations of these notions in terms of averages of classical first-order formulas over suitable probabilistic partitions (Theorems 3.2 and 3.3). We compare these notions with $fim$, $fam$, and self-averaging, and show that for types the notions $fim$, $irgs$, and $rgs$ coincide. We prove that if a definable global Keisler measure$μ$ is not dependent, then every Morley sequence in its canonical random extension $r_μ$ is divergent (Theorem 4.7). Consequently, every $irgs$ measure is dependent (Corollary 4.8) and hence symmetric (Corollary 4.13). Furthermore, we show that for $irgs$ measures the model-theoretic instability events $\mathbf{O}^\varphi$, $\mathbf{I}^\varphi$, and $\mathbf{L}^\varphi$ have $\mathbb{P}_μ$-measure zero (Theorem 5.4). In view of the equivalence of $fim$ and $irgs$ subsequently established in [6], this provides an alternative proof of the corresponding result for $fim$ measures in [9].

comments
35 pages. In this version, some corrections, improvements, and updates have been made. Comments welcome. k.khanaki @ gmail.com
fields
Logic
pdf ↗ page arXiv:2605.15870 ↗

Secure uplink transmission in low-earth orbit (LEO) satellite constellations faces significant challenges due to the time-varying spatial distributions of serving and potential eavesdropping satellites, particularly when their angular separations are small and eavesdropper channel state information (CSI) is imperfect. To address these challenges, this paper investigates fluid antenna (FA)-aided robust secure uplink transmission in LEO satellite constellations. To avoid frequent antenna reconfiguration, the FA positions are optimized over an observation interval and remain fixed throughout the interval, while transmit beamforming dynamically adapts to the time-varying satellite channels. We formulate an average worst-case secrecy rate maximization problem by jointly optimizing FA positions and time-varying beamforming under norm-bounded eavesdropper CSI uncertainties and per-slot secrecy rate constraints. Two new alternating optimization (AO) algorithms are developed to solve the resulting non-convex problem. For robust beamforming optimization, the S-procedure, semidefinite relaxation (SDR), and first-order approximation are employed to construct a tractable convex problem, followed by Gaussian randomization for rank-one recovery. For FA position optimization, we derive concave surrogate functions for successive convex approximation (SCA) and develop a differential evolution (DE)-based search strategy to reduce computational complexity. Numerical results demonstrate that both FA-aided schemes consistently outperform conventional fixed-position antenna arrays across different constellation configurations and CSI uncertainty levels. Moreover, the two algorithms exhibit different performance-complexity tradeoffs as the number of FA elements increases, with DE offering a computational advantage for larger arrays while SCA can achieve higher secrecy rates.

fields
Information Theory
pdf ↗ page arXiv:2605.18099 ↗

Let K be a complete, non-archimedean valued field with a residue field of characteristic different from 2. A Whittaker group G is a discontinuous subgroup of PGL(2,K), freely generated by elements s_0,...,s_g of order two, each defined by a pair of fixed points {a_0,b_0},...,{a_g,b_g}. These fixed points are called ``in good position''. A subgroup W in G of index 2 is a Schottky group and produces a hyperelliptic Mumford curve Omega/W --> Omega/G = P^1, called `Whittaker curve', of genus g and with branch locus B in P^1(K). An explicit parametrization of Whittaker curves in terms of theta functions for W and G and the data of the fixed points, is developed. In particular, this allows one to express the branched points (and other data such as p-adic periods and p-adic heights) in terms of values of theta functions. A central theme of this paper is the relation between the fixed points and the branch locus. For a given configuration (P,m) of $g+1$ pairs of points in P^1, one defines a rigid space Fix_{P,m} of fixed points in good position with that configuration and a rigid space of branched points $ Branch_{P,m} in that configuration. A main result is that the natural morphism FB: Fix_{P,m} --> Branch_{P,m} is a rigid etale covering with Galois group {\pm 1}^{d-1} for some d>0. For all cases of genus g=2,3 (and for some more), an approximation of FB is computed which confirms the main result. Classification of Whittaker groups and analytic reductions of Whittaker curves is another important issue of this paper. The background material in this paper complements the work of L.~Gerritzen, G.~Van Steen, F.~Herrlich and others. It involves re-examination of some proofs, the derivation of properties of semi-stable analytic reductions and studying good position of fixed points.

comments
41 pages
fields
Algebraic Geometry
pdf ↗ page arXiv:2605.22406 ↗

We prove that the topological flow category $\mathcal{M}$ arising from a Morse-Smale pair $(f,ξ)$ on a smooth closed manifold $X$ is equivalent, as an $\infty$-category, to Lurie's $\infty$-category $\mathrm{Sing}_A(X)$ of exit paths in $X$ with respect to the stratification by the stable manifolds of $ξ$. The objects of $\mathcal{M}$ are the critical points of $f$, and for every pair of critical points, the space of morphisms of $\mathcal{M}$ between these is the space of possibly broken trajectories of $ξ$ connecting them; it can be identified up to homotopy with the space of unbroken ones. The latter maps naturally to the space of exit paths connecting these critical points; we prove this map to be a weak homotopy equivalence. Then, we combine these ingredients with several others to construct a zigzag of equivalences between the homotopy coherent nerve of $\mathcal{M}$, denoted $\mathcal{N}(\mathcal{M})$, and $\mathrm{Sing}_A(X)$. The $n$-simplices of $\mathcal{N}(\mathcal{M})$ are homotopy coherent diagrams of $n$ composable morphisms of $\mathcal{M}$; we introduce the notion of unbroken diagram, yielding an $\infty$-subcategory of $\mathcal{N}(\mathcal{M})$, which we refer to as the flow coherent nerve of $\mathcal{M}$. The simplices of the latter give rise to stratified maps out of a family of stratified cubes, into $X$. We organize this family into a functor from the category of finite ordered sequences of critical points, to the category of $A$-stratified topological spaces, and we prove a comparison result with the usual stratified geometric realization functor. We finally use a theorem of Tanaka that associates a functor to a semi-simplicial map between $\infty$-categories that satisfies certain conditions. Our theorem has implications regarding constructible sheaves and the description of homotopy types in terms of flow categories.

comments
173 pages. Minor corrections and clarifications following thesis reports, mainly in Section 3
fields
Algebraic Topology, Category Theory, Symplectic Geometry
pdf ↗ page arXiv:2605.27112 ↗

We prove undecidability for every positive relevant logic extending the system axiomatized by hypothetical syllogism, prefixing, and suffixing and contained in the logic of the semilattice frame $(P_{\mathrm{fin}}(\mathbb{N}), \cup, \varnothing)$. This settles the longstanding decision problem for the semilattice relevant logic S in the negative, contrary to prevailing expectations of decidability. It also provides a new proof of Urquhart's (1984) undecidability theorem for R, E, and T, now by reduction from the Wang tiling problem for arbitrarily large finite isosceles right triangular regions of the plane.

comments
27 pages
fields
Logic
pdf ↗ page arXiv:2605.29880 ↗

Fractional Gaussian noise and $\FARIMA(0,d,0)$ have the same long-memory pole $|θ|^{-2d}$ and hence the same leading PACF law $α(n)\sim d/n$. We show that this agreement breaks at the first non-universal order. For $0<d<1/2$, the pure fGn PACF satisfies $$ α_{\fGn}(n)=\frac d n+\frac{C_{\fGn}(d)}{n^2}+o(n^{-2}), \qquad C_{\fGn}(d)<d^2, $$ The proof uses the Bingham–Inoue–Kasahara representation, a phase-coefficient expansion for fGn, and a Hankel-operator perturbation argument. Thus the fGn spectral envelope is invisible at first order but visible in second-order finite prediction, explaining why short-memory order selection can differ when fGn data are fitted by FARIMA-type models.

comments
There are some mistakes in the manuscript and in real data we can not separate them
fields
Statistics Theory, Probability
pdf ↗ page arXiv:2605.31416 ↗

A point source of light is placed inside a billiard with a smooth, convex, closed boundary. For any integer $n$, the $n$-th caustic by reflection, denoted by $Γ_n$, is the envelope of light rays that have undergone $n$ reflections in such a billiard after emanating from the source. It has been conjectured by Gil Bor and Serge Tabachnikov that for an elliptical billiard, $Γ_n$ has exactly four ordinary cusps; this problem is a billiard variation of Jacobi's Last Geometric Statement, which concerns the number of cusps in the conjugate locus of a point on an ellipsoid. Gil Bor, Mark Spivakovsky, and Serge Tabachnikov have proven that $Γ_n$ has at least four ordinary cusps. In this paper, we present a proof that $Γ_1$ has exactly four ordinary cusps, using billiards in complex spaces.

fields
Dynamical Systems, Algebraic Geometry, Complex Variables
pdf ↗ page arXiv:2606.04132 ↗

We study decentralized stochastic smooth convex optimization, where $M$ workers minimize an average objective using local stochastic gradients and neighbor-only communication over a fixed gossip network. A central question in this setting is to determine the largest number of workers that can be used under a total budget of $N$ gradient samples while still preserving the centralized $O(1/\sqrt N)$ statistical rate. We introduce an accelerated decentralized method that preserves this rate for up to $\smash{M\lesssim \sqrtρ\,N^{3/4}}$ workers, where $ρ$ is the spectral gap of the gossip network, improving the best prior maximal scaling of $\smash{M\lesssim ρ\sqrt N}$. The method is based on a one-step-delayed stochastic acceleration scheme that enables workers to interleave minibatching with accelerated gossip while controlling residual disagreement, and its guarantee depends only logarithmically on the optimum-local heterogeneity. We also establish a matching lower bound for linear-span decentralized first-order methods, showing that the method is optimal up to logarithmic factors.

comments
12 pages, 3 figures, accepted to Neurips 2026
fields
Optimization and Control
pdf ↗ page arXiv:2606.04757 ↗

We consider a periodic scalar reaction-diffusion system over a two-dimensional domain under natural conditions on the reaction function and with initial condition $θ$. We show that on the global attractor $\mathcal A$ of the resulting dynamical system $(u_θ(t):t>0)$, a reverse Poincaré inequality holds true, and that as a consequence the map $θ\mapsto u_θ(t)$ satisfies an $L^2(\Om)$-Lipschitz stability estimate on $\mathcal A$ for any $t>0$ fixed. We then show that statistical recovery of an initial condition $θ$ in the attractor $\mathcal A$, as well as prediction of the states $u_θ$, is possible from discrete measurements of the system at parametric $1/\sqrt N$ convergence rates.

fields
Statistics Theory, Analysis of PDEs, Dynamical Systems
pdf ↗ page arXiv:2606.06018 ↗

In 2025 Cornet, Dravec, and Torres determined the domination number $γ(J(n,3))$ of the Johnson graph for every even $n \ge 6$, expressing it as a closed form $φ_n$ in terms of Fort-Hedlund covering numbers, and conjectured the same value for odd $n$. We prove this conjecture: $γ(J(n,3)) = φ_n$ for every odd $n \ge 7$. A dominating set of $J(n,3)$ is the same as a lotto design $\mathrm{LD}(n,3,3,2)$, and the lotto numbers $L(n,3,3,2)$, and hence $γ(J(n,3))$, were already determined for every $n$ by Bate (1978) and by Brouwer (1981); a short degree argument of Füredi, Székely, and Zubor (1996) gives the same lower bound for all $n \not\equiv 9, 10 \pmod{12}$. The contribution of this paper is therefore a new proof of the odd case, carried out within the bijective framework of Cornet, Dravec, and Torres.

comments
18 pages. v3: adds the lower-bound argument of Füredi, Székely, and Zubor (1996); attributes the extremal structure for odd n >= 13 to Brouwer (1981, Theorem 2); shortens the proof of Lemma 2.4. v2: acknowledges that the result follows from earlier work of Bate (1978) and Brouwer (1981) on lotto designs; corrects several statements about prior work
fields
Combinatorics
pdf ↗ page arXiv:2606.10326 ↗

Let $K_1=B[0,1]$ be the trace of a planar Brownian motion, and let $N_{\mathrm{in}}(η)$ count the bounded components of $\mathbb{R}^2\setminus K_1$ with inradius greater than $η$. As $η\downarrow0$, we prove that $η^2(\log(1/η))^2N_{\mathrm{in}}(η)\to c_{\mathrm{in}}=(π/2)\int G^2d\mathcal{L}\in(0,1/2]$ almost surely and in $L^p$ for every $p\in[1,2]$, where $G$ is the inradius divided by the square root of the area and $\mathcal{L}$ is Werner's limiting shape law. Inradius determines survival under thickening: a cavity contains a hole of the Wiener sausage of radius $r$ if and only if its inradius exceeds $r$. Brownian scaling shows that, for fixed $r>0$ as $T\to\infty$, the number of cavities of $B[0,T]$ with inradius greater than $r$ is asymptotic to $4c_{\mathrm{in}}T/(r^2(\log T)^2)$ in probability and in $L^p$, $p\in[1,2]$. It also gives lower bounds for the sausage's hole count, the Betti curve of its degree-one persistent homology, and its radius integrals, with constant $4c_{\mathrm{in}}$ in expectation and any smaller constant with probability tending to one. A matching upper bound cannot follow from general bounds in terms of perimeter, area, diameter or length. We prove that it is equivalent to a first-moment estimate for near-critical pockets: holes at level $η$ on which the distance to the trace is at most $2η$. Equivalently, it suffices to control residual pockets beyond one per cavity with inradius in $(η,2η]$. This estimate is proved in a companion paper, yielding the exact order $T/(r^2(\log T)^2)$ for the expected number of holes.

comments
This version supersedes the former version v1 (posted under the title "Persistent homology of the planar Wiener sausage: Brownian scaling and a logarithmic expectation law"). It corrects errors in v1 and substantially revises it
fields
Probability
pdf ↗ page arXiv:2606.11248 ↗

For every closed surface of genus at least $3$, equipped with any bounding spin structure, we show that the index difference, viewed as a map from the fundamental group of the space of Dirac-invertible Riemannian metrics to $\KO^{-4}(*)$, is non-trivial. If the genus is at least $5$, we also show that the aforementioned index difference is surjective. For products of two closed surfaces of genus at least $3$, equipped with any spin structure, we prove that the corresponding space of Dirac-invertible Riemannian metrics is not contractible. We discuss the relationship of this result to the existence of metrics with harmonic spinors in dimension~$4$.

comments
Minor changes
fields
Differential Geometry, K-Theory and Homology
pdf ↗ page arXiv:2606.12283 ↗

For a polarized complex manifold with discrete automorphism group, it is known that if the first Chern class admits a cscK metric, then the balanced metrics, which are characterized in terms of the algebro-geometric notion of Chow stability, approximate this cscK metric. In this paper, we study a non-Archimedean analogue of this phenomenon. In particular, we prove that such an analogue holds for polarized totally degenerate abelian varieties. As an application, we also show that, for a totally degenerating family of polarized abelian varieties, the validity of this non-Archimedean analogue yields a uniform estimate for the Calabi–Yau metrics on fibers sufficiently close to the degenerate fiber.

comments
35 pages. Comments are welcome! v3: Corrected errors and improved the exposition. In particular, Theorem A, which was strengthened in v2, has been reverted to its original form in v1. All results stated in v1 remain valid
fields
Algebraic Geometry, Differential Geometry, Number Theory
pdf ↗ page arXiv:2606.15758 ↗

An equivalent version of the Báez-Duarte criterion \cite{baez} for the Riemann Hypothesis (RH) by Bagchi states that the RH holds true if and only if the function $E(s) = 1/s$ belongs to the closed linear span of $G_k(s) = (k^{-s} - k^{-1})ζ(s)/s,\, k \geq 2$ in the Hardy space \( H^2(\mathbb{C}_{1/2}) \), where $\mathbb{C}_α$ denotes the half-plane $\mathrm{Re}(s)>α$. We first show that if $E$ belongs to the closure of span$(G_k)_{k\geq 2}$ in \( H^2(\mathbb{C}_α) \) for $α>1/2$, then $ζ$ is zero-free in $\mathbb{C}_α$. We then use this as the basis for a numerical analysis of the sequence \[ s_n = \left\| \sum_{k=2}^{n} μ(k) G_k - E \right\|^2_α, \] for $1/2\leq α\leq 1$, where $\left\|.\right\|_α$ is the norm in $H^2(\mathbb{C}_α)$ and $μ$ the Möbius function.

comments
This article is withdrawn by the authors. It is superseded by arXiv:2611.03924, a separate submission with a revised author list reflecting the current composition of the research team. Readers should refer to arXiv:2611.03924
fields
Functional Analysis, Complex Variables
pdf ↗ page arXiv:2606.16097 ↗

We develop a new theory of thermodynamic formalism for non-compact flows, providing criteria for existence and uniqueness of equilibrium states, as well as ergodicity and a Gibbs property. These criteria include new versions of the expansivity and specification properties that hold for geodesic flows on non-compact negatively curved manifolds. Previous results on thermodynamic formalism for these geodesic flows were restricted by a pinched curvature condition, which we remove, allowing curvature to approach both $0$ and $-\infty$. Our results also apply to geodesic flows beyond the smooth setting, including locally CAT(-1) spaces.

comments
v3: 93 pages. Improved introduction
fields
Dynamical Systems
pdf ↗ page arXiv:2606.20405 ↗

The Proca equations (1936) are used in quantum field theory to describe vector bosons (spin 1) with a nonzero mass. The Proca equations are not gauge invariant. In contrast to Stueckelberg's approach (1938), this article presents a gauge invariant generalization of the Proca equation by introducing an additional vector field into the Proca equation. The results are extended to the Yang-Mills-Proca equations, leading to equations with non-Abelian gauge symmetry.

comments
10 pages
fields
Mathematical Physics
pdf ↗ page arXiv:2606.20828 ↗

In this paper, we provide sufficient conditions for the absolute continuity of one-dimensional push-forwards of dependent random vectors. Suppose that $X$ has an absolutely continuous distribution and that the conditional distribution of an $\mathbb{R}^d$-valued random vector $Y$ given $X=x$ is nondecreasing in $x\in \mathbb{R}$ in the usual stochastic order. For Borel maps $g\colon \mathbb{R}\times\mathbb{R}^d\to\mathbb{R}$ satisfying a coordinatewise monotonicity condition in $Y$ and a uniform lower-increment condition in $X$, we prove that $g(X,Y)$ has an absolutely continuous distribution. The result requires neither independence nor a joint density, and allows the marginal law of $Y$ to be completely arbitrary. Moreover, the result remains valid if $\mathbb{R}^d$ is replaced by an arbitrary measurable space endowed with a reflexive binary relation. We discuss consequences for monotone risk aggregation and extensions of the familiar regularization by convolution beyond independent random variables.

fields
Probability
pdf ↗ page arXiv:2606.21042 ↗

We consider localized oscillations of an Euler–Bernoulli beam on a visco-elastic foundation coupled to a damped discrete oscillator. All system parameters vary slowly and independently in time. For the conservative case, we employ three different analytical approaches: an asymptotic method, a technique based on the adiabatic invariance of the action of a trapped wave, and an analysis of the equivalent Hamiltonian system. All three approaches yield the identical formula for the oscillation amplitude. In the dissipative case, the oscillation amplitude is obtained solely via the asymptotic approach. Finally, we validate our analytical findings against numerical simulations, demonstrating excellent agreement.

comments
15 pages, 1 figure
fields
Mathematical Physics
pdf ↗ page arXiv:2606.22444 ↗

In this article, we study Riemannian planes $(M,g)$ which satisfies a bounded geometry condition ( bounded sectional curvature and a non-vanishing injectivity radius of $(M,g)$). We show that such a Riemannian plane is always quasi-isometric to a radially symmetric plane which also satisfies the aforesaid bounded geometry condition. Using that, we study the quasi-redirecting boundary of Riemannian planes with bounded geometry and show that it is either a single point or $\mathbb{S}^1$. We contrast the quasi-redirecting boundary with the Martin boundary and construct Riemannian planes with bounded geometry for which the We also characterise quasi-redirection in Riemannian planes with bounded geometry via their asymptotic cones and their cut points. Lastly, we characterise $\mathbb{E}^2$ and $\mathbb{H}^2$ upto quasi-isometry by their asymptotic cones.

comments
22 pages, 0 figures
fields
Differential Geometry
pdf ↗ page arXiv:2606.27393 ↗

We introduce a new elementary technique for proving $\#P$-hardness results under Turing reductions, and showcase its applicability by settling the counting complexity of several fundamental problems about counting orderings in graphs, posets, and geometric configurations. The technique addresses a natural difficulty in counting reductions: when designing gadgets that attempt to enforce a semantic property, the introduction of the gadgets themselves can affect the function to count in complicated ways. In a nutshell, we introduce an amplification parameter $q$ for the gadgets (e.g., replacing a subset of the vertices of a graph by $q$ copies) and show that the limiting behavior when $q$ tends to infinity, or even to negative values, filters out the unwanted effects of the gadgets, which allows us to recover the desired counts. These limits can be computed symbolically via interpolation, after showing that the probability of a random ordering satisfying the problem's constraints is a rational function in $q$. Concretely, with this methodology, we show $\#P$-completeness for counting the following objects: (a) successive vertex orderings, even on bipartite graphs, (b) $st$-numberings of graphs, (c) shellings, even for bipartite graphs, (d) perfect elimination orderings, even for split graphs, (e) linear extensions of N-free posets of height $3$, (f) linear extensions of posets of height $2$, and (g) peeling sequences of point sets. Result (e) settles a conjecture of Felsner and Manneville (2015), and although result (f) was first proved by Dittmer and Pak (2018), we include an alternative proof, using our technique, that does not rely on the result of Brightwell and Winkler (1991) about hardness of counting for general posets. Result (g) answers a question of Eppstein.

fields
Combinatorics
pdf ↗ page arXiv:2606.29157 ↗

Many measurements in physics determine a linear transformation from pairs of input and output vectors, where the transformation is known in advance to belong to a matrix Lie group, for example the Lorentz group for the four-momenta measured by two inertial observers, or the unitary group for the states produced by a quantum circuit. Among the classical groups, only the rotation and unitary groups have a well-known closed form solution, given by the Kabsch and Horn algorithms and by the polar decomposition. We extend our previously proposed Lie algebra method from the image of the matrix exponential to the entire Lie group. We derive the projection in closed form for the classical groups over the real and complex numbers, including the indefinite orthogonal and unitary, symplectic, spin, special Euclidean and similarity groups, and we give conditions under which the method is exact for noiseless data, with examples in which it fails otherwise and an extension by a data dependent involution which is exact on the whole group. For noisy data, we bound the error of the method to first order in the noise, and we show that, for the orthogonal and unitary groups, it agrees with the least squares solution over the group to first order when the input vectors are isotropic, but not otherwise. We apply the method to an eight-qubit quantum circuit and to the rigid registration of point clouds, and compare it with Gauss–Newton refinement and direct minimization. A Julia implementation is provided.

comments
44 pages
fields
Numerical Analysis, Mathematical Physics
pdf ↗ page arXiv:2606.30868 ↗

We express the log discrepancies of co-rank one foliations on normal varieties in terms of arc codimensions and differential orders. On smooth threefolds, if a foliation has simple singularities and its separatrices form an algebraic simple normal crossings divisor, we identify its foliated discrepancies with those of an ordinary pair for every divisor centred on the separatrices. This gives adjunction and log canonical threshold formulae. For algebraic images of the separatrices, we estimate changes in boundary orders caused by failure of ideal generation and by descent through the conductor. For locally complete intersection images, comparison on a common resolution yields criteria for equality of discrepancies and bounds on log canonical thresholds.

comments
27 pages, major revision with improved results
fields
Algebraic Geometry, Complex Variables, Dynamical Systems
pdf ↗ page arXiv:2607.01809 ↗

We exhibit a formula relating the correlation between Wilson loop variables in Potts lattice gauge theory to a topological quantity in the plaquette random cluster model. As applications we prove exponential decay of correlations between slowly growing Wilson loop variables for Ising lattice gauge theory on $\mathbb{Z}^3$ at all but the critical temperature, we demonstrate that the correlation length is finite at sufficiently high or low temperatures in any dimension, and we show that the correlation length of the model on $\mathbb{Z}^4$ with free boundary conditions equals that of the dual model with constant boundary conditions when $q$ is prime.

comments
Significant revisions, including corrections of errors in the proofs of Theorems 8 and 9. Added an appendix with an AI-assisted proof restoring Theorem 9 to its original generality. The main text of the paper remains human-generated
fields
Probability, Mathematical Physics, Algebraic Topology
pdf ↗ page arXiv:2607.02434 ↗

We study an extension of the algebra of the infrared to curve–valued potentials, focusing on the elliptic curve case. Given a finite configuration of points on an elliptic curve, we construct associated \(L_\infty\)-algebra. Adding an extra point produces a directed \(A_\infty\)-algebra and an \(L_\infty\)-morphism to their directed Hochschild deformation complexes. Maurer–Cartan elements of the \(L_\infty\)-algebra determine \(A_\infty\)-deformations of the directed algebra. We formulate a conjectural comparison between Maurer–Cartan deformations of the directed algebra and the directed Fukaya–Seidel category associated to compatible finite collections of thimbles.

comments
60 pages, 5 figures
fields
Symplectic Geometry, Quantum Algebra
pdf ↗ page arXiv:2607.04039 ↗

We introduce the microscopic weighting, a canonical signed measure of mass one that can be associated to almost any finite metric space. The microscopic weighting is obtained as the small-scale limit of the weightings used to define the magnitude function. We give general criteria for its existence, proving in particular that every finite space of strictly negative type admits a microscopic weighting; this includes every finite subset of Euclidean or hyperbolic space and every finite tree. Heuristically speaking, the microscopic weighting distributes its mass as widely as possible across a space, assigning greater weight to sparse or outlying regions and emphasizing points on the boundary. Indeed, we show that on a finite space of negative type the microscopic weighting can be characterized (when it exists) as an optimizing measure for an energy integral determined by the distance function. Alternatively, it can be characterized in terms of the geometry of the Schoenberg embedding. Each of these interpretations also clarifies the information carried by the derivative of the magnitude function at zero. Though our main focus in this paper is on finite metric spaces, we lay the groundwork to extend the theory to compact subsets of Euclidean space. In that setting, we observe that the microscopic weighting must be understood as a distribution rather than as a measure.

comments
38 pages, 4 figures. Changes in V2: We include a stronger existence theorem (Theorem 3.7). Section 3 has been rewritten to accommodate it, and the proof is given in the new Section 7. Conjecture 3.3 of V1 has been disproved; see Remark 3.3 of V2. Remarks 5.7 and 5.14 have been added, and with them several new references
fields
Metric Geometry, Combinatorics
pdf ↗ page arXiv:2607.05349 ↗

Dimensionality reduction has proven powerful for identifying neural manifolds, which are low-dimensional structures underlying high-dimensional neural activity. These low-dimensional representations have improved the interpretability of population-level coding. Yet whether such low-dimensional representations are biologically relevant and confer functional advantages in learning systems, or merely reflect neuron-level activity, remains contested in neuroscience. We show that an explicit information bottleneck forcing a recurrent neural network to learn a low-dimensional representation is necessary for rotational and out-of-distribution generalisation in a time-series prediction task. Using information-theoretic measures of causal emergence, we characterise the dynamics of this representation across the memorisation-to-generalisation transition, finding a non-monotonic trajectory which shows an initial decrease, a minimum, and a subsequent rise to a maximum, even as prediction loss falls monotonically. This trajectory scales with task complexity, and the magnitude of emergent structure reliably predicts generalisation performance. Analysis of hippocampal CA1 and medial prefrontal activity in rats learning an alternating maze task reveals analogous non-monotonic emergence dynamics that track behavioural performance. Together, these findings indicate that the ability of neural networks to learn compact, distributed and emergent representations confers a functional advantage for generalisation, supporting a causal role for learned representations in cognition.

fields
Information Theory
pdf ↗ page arXiv:2607.10430 ↗

This paper presents a symplectic Hamiltonian direct discontinuous Galerkin (DDG) method for approximating wave propagation problems, including the linear and semilinear wave equations. Within an auxiliary-variable-free DG framework, we prove that the symmetry of the numerical flux bilinear form is equivalent to the existence of a discrete Hamiltonian structure. It follows that methods such as the symmetric interior penalty, BR2, and symmetric DDG (SDDG) methods admit a discrete Hamiltonian structure, whereas the Baumann–Oden and DDG methods do not possess this property. Exploiting this structure, we construct fully discrete symplectic schemes by combining the SDDG spatial discretization with symplectic time integrators. We also develop a unified treatment of periodic, Dirichlet, Neumann, and Robin boundary conditions that preserves the discrete Hamiltonian structure of the SDDG scheme. We further derive error estimates for the SDDG method applied to semilinear wave equations under these boundary conditions, showing the optimal convergence rate for the displacement and the suboptimal convergence rate for the velocity. Numerical experiments validate the theoretical convergence rates and demonstrate that the symplectic Hamiltonian DDG method achieves superior long-time energy conservation and accuracy.

comments
43 pages, 11 figures
fields
Numerical Analysis
pdf ↗ page arXiv:2607.10652 ↗

Hales-Jewett theorem is well known. We prove a primitive recursive bound, in fact an $\clE^5$ one, for the density version. In preparation we deal with what was the main induction step in the old proof of Hales-Jewett as a result in its own right with better bounds. Also in improving our bounds.

comments
proofread and expanded
fields
Combinatorics
pdf ↗ page arXiv:2607.14732 ↗

We study stochastic fast-slow systems with noise acting exclusively on the slow variable near uniformly attracting, normally hyperbolic critical manifolds $x=X^\ast(y)$. The deviation equation contains transmitted diffusion and an Itô curvature forcing of order $O(σ^2\lVert D^2X^\ast\rVert)$. Using a parameter-dependent Lyapunov function and stopped martingale estimates, we obtain geometry-dependent concentration bounds under the sufficient noise restriction $σ\le C_0\min(\sqrt{\varepsilon},\varepsilon^{1/4}L_{\mathrm{geom}}/\sqrt{λ_0})$, where $λ_0$ is the uniform spectral gap and $L_{\mathrm{geom}}=\sqrt{λ_0/(\lVert D^2X^\ast\rVert\lVert h\rVert^2)}$. For fixed geometry and any fixed $p>0$, the confidence radius on $[0,T/\varepsilon]$ is $O(\varepsilon+σ^2+σ\sqrt{\log(1/\varepsilon)})$, with stopped fast-variable failure probability at most $\varepsilon^p$; slow-variable exit is controlled separately. We also establish an adiabatic error bound of order $O(\varepsilon+σ\sqrt{\varepsilon}+σ^2\lVert D^2X^\ast\rVert)$, with explicit failure probabilities. These estimates quantify the curvature dependence of fast-variable concentration and its $O(\varepsilon)$ contribution to the slow-error upper bound for fixed geometry.

comments
40 pages, 1 figure. Revised version with corrected long-time probability estimates and updated confidence-radius bounds; proofs and exposition revised
fields
Probability, Dynamical Systems
pdf ↗ page arXiv:2607.16217 ↗

We extend the definitions of kappa, psi and lambda classes from the case of moduli spaces of stable curves to the higher-dimensional case of KSBA spaces, and prove a number of their basic properties.

comments
v2: Title change. Added psi and mixed classes and computation for a series of Calabi-Yau moduli
fields
Algebraic Geometry
pdf ↗ page arXiv:2607.19251 ↗

The Ulam–Kaplansky problem up to similarity in operator algebras asks whether $A\oplus A\sim B\oplus B$ implies $A\sim B$. We construct block-diagonal compact operators for which this implication fails. For von Neumann algebras with separable predual, we prove that this cancellation property is equivalent to subhomogeneity. Moreover, we obtain a cancellation law for finite ampliations in unital Banach algebras when the commutant of one element is finite dimensional modulo its Jacobson radical.

comments
21 pages
fields
Functional Analysis, Operator Algebras
pdf ↗ page arXiv:2607.20593 ↗

Let $G$ be a graph with adjacency matrix $A$ and degree matrix $D$. Applying $A$ and $D$ repeatedly to the all-ones vector produces a space $M_G$ of functions on the vertices, and these functions are constant on the orbits of the automorphism group. We call $G$ AD-observable if $M_G$ contains the indicator vector of every orbit, and AD-separable if the functions in $M_G$ separate the orbits. Forests are AD-observable, by McKay's leaf-pruning argument. Our main result is that for every cograph, $M_G$ is spanned by the indicator vectors of the cells of the stable partition of color refinement. Hence a cograph is AD-observable, or equivalently AD-separable, exactly when color refinement finds its orbits, and we describe these cographs by their cotrees. We also show that two cographs with the same number of vertices and the same homomorphism counts from all caterpillars have the same counts from all trees. The main result does not extend to distance-hereditary graphs: we give two such graphs on nine vertices that are AD-separable but not AD-observable.

comments
The main results are reformulated from v1 and results on cographs are added
fields
Combinatorics, Spectral Theory
pdf ↗ page arXiv:2607.21494 ↗

We use a gluing procedure introduced by Ginzburg to describe the relative Langlands dual of the hyperspherical Hamiltonian $(\mathrm{GL}_n \times \mathrm{GL}_m)$-variety $T^*(\mathrm{Hom}(\mathbb{C}^m,\mathbb{C}^n))$, and in particular the Rankin-Selberg case $m=n$. We show that the dual is isomorphic to the triangle part of Cherkis-Nakajima-Takayama bow varieties, recovering a result of Nakajima. Using this result, we give a new proof of a theorem of Braverman-Finkelberg-Nakajima describing the regular perverse sheaf in type $A$. Following a suggestion of Ginzburg, we explain how to modify the gluing so that the dual Hamiltonian variety of $T^*\mathbf{N}$, for any gluable finite-dimensional representation $\mathbf{N}$ of a complex reductive group $G$, is naturally equipped with an anti-symplectic involution, and give an explicit formula for this involution in the Rankin-Selberg case.

comments
20 pages, v2: added section 5
fields
Representation Theory
pdf ↗ page arXiv:2607.22965 ↗

We give a direct differential-geometric proof of a result of Crainic on the cohomology of pullbacks of Lie algebroids under surjective submersions: If the fibers are connected and have vanishing cohomology through degree $n$, pullback with coefficients in any representation is an isomorphism through degree $n$ and is injective in degree $n+1$. The proof reduces to a vanishing result for leafwise cohomology and uses an exhaustion-and-patching argument due to Buchdahl. No local triviality or finite-type assumption on the fibers is required. We also prove the analogous result for holomorphic Lie algebroids.

journal
SIGMA 22 (2026), 107, 13 pages
fields
Differential Geometry
pdf ↗ page arXiv:2607.23859 ↗

The Jacobian conjecture, open since 1939, asks whether every polynomial map of $\C^n$ whose Jacobian determinant is a nonzero constant must have a polynomial inverse. It was refuted in dimension three by Alpöge on July 19, 2026, with an infinite family by Gallagher (July 20) and a geometric explanation by Speyer (July 23): the counterexample sweeps the tangent lines of a plane curve. We give a self-contained account of this tangent-sweep mechanism and generalize it from plane curves to direction fields on hypersurfaces. The resulting construction, including stabilization of the three-dimensional family, produces counterexamples in every dimension greater than two and, in each dimension, of arbitrarily large geometric degree (the number of preimages of a typical point). We work it out in five explicit maps: one three-dimensional of degree four, two four-dimensional of degrees five and ten, and two five-dimensional of degrees six and twelve. We show that our explicit examples in dimensions $n\ge4$ are not polynomially left-right equivalent to the counterexamples of Alpöge and Gallagher when these are embedded in $\C^n$ by adjoining identity coordinates. The counterexamples provide explicit examples of étale morphisms $\C^n\to\C^n$ that are not proper: they are everywhere unramified, and fail to be injective only through points escaping to infinity. All identities were verified in exact rational arithmetic. Part~\ref{part:equivalence} proves the higher-dimensional inequivalence results, and Appendix A determines exact fiber structures for the three-dimensional examples. A complete classification of the fibers of $F_4,F_5,F_6,F_7$ remains open.

fields
Algebraic Geometry
pdf ↗ page arXiv:2608.00222 ↗

Total quadratic ball discrepancy measures the deviation of codeword counts from uniformity over all Hamming balls. We prove that, whenever a Hamming space and cardinality admit a perfect code, the discrepancy minimizers among all subsets of that cardinality are precisely the perfect codes. This extends Barg's binary minimizing result and characterizes equality over arbitrary finite alphabets. At parameter sets that satisfy the sphere-packing condition and Lloyd's integral-root condition for correcting one or two errors, we give an explicit lower bound whose attainment is equivalent to perfect tiling: the discrepancy that a perfect code with these parameters would have. For one error, these include parameters at which the existence of a perfect code is open. For alphabets of size at least four, the discrepancy is an affine function of an energy with a strictly completely monotonic potential; for these alphabets, known universal energy bounds also give these lower bounds, and, when a perfect code exists, known universal optimality results also give the characterization of the minimizers. Over the ternary alphabet, complete monotonicity fails, and we prove the ternary bounds directly at all Hamming parameters and at the Golay parameters, the only nontrivial ternary cases.

comments
38 pages; no figures. Revised and expanded presentation, clarified hypotheses and supporting arguments, and updated references. Exact-arithmetic verification companion: https://doi.org/10.5281/zenodo.22084106
fields
Information Theory, Combinatorics
pdf ↗ page arXiv:2608.01134 ↗

Unbiased squared distance covariance and the Hilbert-Schmidt independence criterion (HSIC) are fourth-order U-statistics, yet U-centering evaluates them from pairwise arrays in $O(n^2)$ operations. We show that U-centering is exactly least-squares residualization after fitting additive endpoint effects. This analysis of variance (ANOVA) interpretation explains the zero row sums and the normalization through residual degrees of freedom. The same geometry gives regression representations of dependence and two-sample statistics, exact permutation variances, and relations among pooled and regression mean squares. For arrays indexed by $r$-element subsets and $n\ge2r$, higher-order U-centering removes all lower-order subset effects and projects onto a space of dimension $\binom{n}{r}-\binom{n}{r-1}$. For fixed symmetric square-integrable kernels evaluated on independent and identically distributed (i.i.d.) paired observations, we derive the residual covariance matrix, exact mean-squared errors for estimating the highest-order Hoeffding components, unbiased positive semidefinite (PSD) covariance estimates, and moment formulas under independence and permutation. The normalized residual inner product equals a direct unbiased estimator involving up to $2r$ observations, but can be evaluated in $O(n^r)$ operations for fixed $r$. For four-argument matching kernels, three matrix products yield exactly unbiased estimators of a fourth spectral moment of kernel cross-covariance in $O(n^3)$ arithmetic and $O(n^2)$ storage. A numerical illustration distinguishes dependence structures with equal HSIC and different spectral concentration.

comments
57 pages, 1 figure. Expanded covariance, component estimation and permutation results; added unbiased cubic-time spectral estimation and a numerical illustration
fields
Statistics Theory
pdf ↗ page arXiv:2608.01364 ↗

We study the BRST quantization of the null string after completing its local gauge symmetry by Carroll-Weyl transformations. The resulting worldsheet theory possesses three first-class constraints, $C_1 = P^2$, $C_2 = P \cdot X'$, and $C_3 = P \cdot X$, whose modes realize a Weyl-BMS algebra. The additional Carroll-Weyl constraint qualitatively changes the quantum gauge complex: its scalar $s$-ghost is intrinsically coupled to the BMS $bc$-ghost sector, and the anomaly analysis involves three independent cocycles rather than a single Virasoro-type central charge. Starting from the gauge-fixed action, we derive the complete Faddeev-Popov complex, construct the matter and ghost currents and the BRST charge, and evaluate their equal-time operator products in the flipped, equivalently highest-weight, representation. The matter and ghost anomaly coefficients are $(c_{LL},c_{LS},c_{SS})_{\mathrm{matter}}=(2D,-D,-D)$ and $(c_{LL},c_{LS},c_{SS})_{\mathrm{ghost}}=(-54,6,4)$. Because the corresponding central terms multiply linearly independent ghost bilinears in $Q_B^2$, BRST nilpotency requires the three conditions $D=27$, $D=6$, and $D=4$, respectively. These conditions are mutually incompatible. Consequently, there is no target-space dimension in which the minimal flat Carroll-Weyl matter-plus-ghost complex is anomaly-free in the highest-weight representation. The familiar $D=26$ condition of the ILST null string is recovered only after truncation to the two-constraint BMS subsector, which defines a different quantum gauge complex.

comments
40+17 pages
fields
Mathematical Physics
pdf ↗ page arXiv:2608.02731 ↗

A class of elliptic solutions to the matrix CKP equation is studied. Equations of motion for their poles and matrix coefficients at the poles are obtained. As in the scalar case, they are of the first order, in contrast to what takes place in the KP and BKP hierarchies, where the equations of motion are of the second order.

comments
15 pages, no figures. Presentation improved, typos corrected
fields
Mathematical Physics
pdf ↗ page arXiv:2608.08774 ↗

Let $W$ be a henselian valuation ring with fraction field $L$, residue field $k$, and value group $Γ_W$. Let $N=\ell^ν$ be invertible in $W$. Choose an ordered $\mathbf Z/N$-basis $B$ of $Γ_W/NΓ_W$. Products of suitable classes define an equivalence of complete filtered spectra $$ \bigoplus_{\substack{J\subseteq B\\J\text{ finite}}} Σ^{|J|}\operatorname{Fil}_{\mathrm{mot}}^{\bullet-|J|} K(k;\mathbf Z/N) \xrightarrow{\simeq} \operatorname{Fil}_{\mathrm{mot}}^{\bullet} K(L;\mathbf Z/N) $$ The summand indexed by $\varnothing$ is the generic restriction map, after the rigidity equivalence $K(W;\mathbf Z/N)\simeq K(k;\mathbf Z/N)$, and is therefore split injective. Independently of this splitting, excision yields a lifting theorem for regular henselian pairs. In particular, this yields finite-coefficient Gersten injectivity for noetherian henselian regular local rings. Applications to completions along regular primes give relative and sometimes nonhenselian examples. More generally, if $P$ is a Prüfer ring and $R$ is a henselian local ind-smooth $P$-algebra, then $R$ is a domain and $K_n(R;\mathbf Z/N)\to K_n(\operatorname{Frac}(R);\mathbf Z/N)$ is injective for every $n$, provided $N\in R^\times$. These injectivity consequences extend from prime-power to arbitrary finite invertible coefficients by primary decomposition.

comments
44 pages, minor changes
fields
K-Theory and Homology, Commutative Algebra, Algebraic Geometry
pdf ↗ page arXiv:2608.18789 ↗

We propose a fully discrete method for linear fourth-order parabolic equations with Dirichlet boundary conditions, combining an implicit LBRFD multistep scheme in time with a mixed interior penalty discontinuous Galerkin (IPDG) method in space. The temporal discretization employs equispaced linear barycentric rational interpolants and incorporates a startup procedure. To facilitate the spatial discretization, the original problem is reformulated through an auxiliary variable. For certain parameter pairs $(n,d)$, the LBRFD method is shown to be $A(α)$-stable and to possess a wider stability angle than the corresponding BDF$p$ method of the same order. Stability and a priori error estimates are established via a $G$-energy technique and the discrete Grönwall lemma. The theoretical analysis yields a total $L^2$ error estimate of order $h^{k-1}+τ^p$, where $p=d$ if $n-d$ is even and $p=d+1$ if $n-d$ is odd. The reduced spatial convergence rate is attributed to boundary contributions on $\partialΩ$. Despite this theoretical prediction, numerical experiments confirm the stability and demonstrate optimal convergence of order $h^{k+1}+τ^p$.

comments
28 pages, 10 figures
fields
Numerical Analysis
pdf ↗ page arXiv:2608.19250 ↗

During clinical trials evaluating a drug's effect on a survival endpoint, intermediate events often occur before the primary event. The treatment can affect the primary event through multiple pathways via intermediate events. With multiple intermediate events, each is a treatment-induced confounder of transitions out of the others, so we cannot identify natural or separable effects. We instead consider stochastic interventions that assign each transition the identifiable observed-history hazard of either arm. Contrasts between the resulting counterfactual cumulative incidences define transition-, event-, and path-targeted effects, identifiable without cross-world or isolation assumptions. Based on the efficient influence functions for the counterfactual cumulative incidences under hypothetical interventions, we construct multiply robust and semiparametrically efficient estimators for these targeted treatment effects. Our proposed framework quantifies treatment effects through each transition, on each event, and along each path. By analyzing data from the LEADER trial, we find that liraglutide significantly reduces the risk of cardiovascular and microvascular events. The reduction in all-cause mortality is primarily mediated by its effects on expanded major adverse cardiovascular events.

fields
Statistics Theory
pdf ↗ page arXiv:2608.22608 ↗

Quadratic ball discrepancy measures how uniformly a code meets Hamming balls over all centers and radii. Perfect codes minimize this quantity among codes of the same cardinality. We prove a quantitative refinement: at the parameters of every perfect code other than a one-word code or the full space, the discrepancy excess over the minimum is at least an explicit positive multiple of the discrepancy at the packing radius. The latter equals the tiling defect divided by the square of the code cardinality. This defect is the total squared deviation of the covering multiplicity from one. The comparison applies to all codes of that cardinality. For one-error perfect codes, the coefficient is at least one, and no larger constant holds for all such codes; explicit coefficients depending on the parameters can be much larger. The tiling defect also measures the failure of uniform ball noise to smooth a code distribution. Consequently, the same discrepancy excess gives explicit bounds on holes, overlaps, and divergence from the uniform distribution.

comments
39 pages; no figures. Revised exposition and reference locators. Exact-arithmetic verification companion: https://doi.org/10.5281/zenodo.22109975
fields
Information Theory, Combinatorics
pdf ↗ page arXiv:2608.25716 ↗

We investigate inverse scattering at a fixed energy for semiclassical Schrödinger operators with smooth potentials. For compactly supported potentials satisfying a natural virial condition, we prove that, under the sharp condition, $\liminf_{h\to0}\|S_1(λ,h)-S_2(λ,h)\|<2$ implies $V_1=V_2$, provided the fixed energy $λ$ lies above both potentials. We then consider radial short-range repulsive potentials. Under a monotonicity assumption on the radial force, we show that the semiclassical differential cross section at a fixed energy, up to $o(1)$ as $h\to0$, determines the potential in the classically accessible region. Finally, we obtain an analogous rigidity result for nontrapping compactly supported perturbations of the Euclidean metric under a strict convexity assumption.

comments
The upper bound constant \(\sqrt{2}\) has been replaced by the optimal constant \(2\)
fields
Analysis of PDEs, Mathematical Physics, Spectral Theory
pdf ↗ page arXiv:2609.02175 ↗

In this paper, we study the emergent collective dynamics of multi-agent systems driven by infinite distributed fading memory of Volterra type. We establish a unified theoretical framework covering both first-order opinion consensus dynamics and second-order velocity alignment flocking kinematics. By introducing Dafermos past-history transformations, the governing integro-differential systems are reformulated into dynamical systems on extended product Hilbert spaces. For first-order dynamics, we prove that fading memory inherently provides a hidden dissipative mechanism, guaranteeing unconditional global exponential consensus with or without instantaneous communication forces. For second-order dynamics, by developing a unified energy-multiplier framework that seamlessly couples instantaneous communication with memory dissipation, we establish unconditional global exponential flocking for both the pure fading memory system and the mixed interaction system, completely independent of the spatial dispersion or the communication tail.

fields
Dynamical Systems, Optimization and Control
pdf ↗ page arXiv:2609.02454 ↗

This paper addresses safe coordination of scalar agents with constrained input dynamics over fixed rooted digraphs. We propose a networked Admissibility-Preserving Control (APC) architecture for a common moving output corridor under heterogeneous asymmetric input bounds, wherein the Admissibility-Preserving Input Realization (APIR) generates the physical inputs. In logarithmic barrier coordinates, the closed-loop dynamics form an exact cascade of exponentially decaying realization errors and directed consensus dynamics. We derive direction-specific compatibility conditions that guarantee uniform corridor and input margins, bounded commands, unique complete solutions, and exponential consensus for compact initial sets. The closed-form expression for the agreement value quantifies the effect of the initial APIR mismatch. Under the corresponding compatibility conditions, partial pinning guarantees safe reference tracking whenever the root component contains at least one informed agent. Simulations illustrate the coordination guarantees.

fields
Dynamical Systems
pdf ↗ page arXiv:2609.09384 ↗

We study fluctuation limit of a slow-fast system driven by $α$-stable Lévy noise with $1<α<2.$ The slow component is generated by an odd polynomial function $f(y):=y^q,$ while in the fast component, the drift is $g(y):=-|y|^p\operatorname{sgn}(y)$ for some $p>0.$ Although the noise is given, the fluctuation limit is either a stable process or a Brownian motion, depending on both $p$ and $α.$ The critical line between stable limit and Brownian limit is $q+1-p=α/2.$

comments
46 pages, 1 figure
fields
Probability
pdf ↗ page arXiv:2609.09906 ↗

Moment Lagrangian $L_μ$ is a Lagrangian in $T^*G^- \times Y^- \times Y$ associated to a Hamiltonian $G$-space $Y$ with a moment map $μ$. In this paper, we prove that $L_μ$ is tautologically unobstructed under mild assumptions on $Y$. As a key ingredient in the proof, we constructed a new symplectic groupoid structure on $T^*G^- \times Y^- \times Y$ over $G\times Y$ for which $L_μ$ is simultaneously the unit and the fixed locus of the inversion, which might be of independent interest.

comments
Revisions on Corollary 1.4 (2): spin assumption added; definition of the character Lagrangian corrected; new remark 1.6 on its context added. Minor revisions elsewhere
fields
Symplectic Geometry
pdf ↗ page arXiv:2609.12033 ↗

Every compact subspace $K$ of a homogeneous generalized ordered (GO) space $X$ satisfies $χ(K)\leqω_1$ and $|K|\leq 2^{ω_1}$. If $χ(K)>ω$ for some such $K$, then $χ(X)=ω_1$ and $πχ(X)=ω$. We construct a homogeneous $σ$-compact linearly ordered topological space (LOTS) $\mathbf{H}$ with a compact subspace $K$ such that $χ(x,K)=ω_1$ for every $x\in K$ and $c(K)=w(K)=|K|=2^{ω_1}$, where $c(K)$ denotes cellularity. Thus both bounds are attained by the same compact space. Every semitopological group that is a GO space is hereditarily paracompact and has metrizable compact subspaces. If it is not a $P$-space, it is also first countable and submetrizable. A semitopological group that is a LOTS, or a topological group that is a GO space, is either metrizable or is a $P$-space. We also show that in an order-homogeneous LOTS every compact subset is first countable.

comments
36 pages
fields
General Topology
pdf ↗ page arXiv:2609.15499 ↗

The existence and structure of non-simple blow-up are fundamental questions for the self-dual Chern–Simons–Higgs equation. Physically, this phenomenon describes magnetic-flux concentration in several cores whose mutual distances are much larger than their widths, although all cores approach the same vortex point. Despite substantial progress for singular Liouville equations, the existence of non-simple blow-up solutions to the full Chern–Simons–Higgs equation on closed surfaces has remained open in both the Chern–Simons (CS) and mean-field regimes. We prove that the cores of every non-simple cluster have equal limiting masses, share a common profile type, and form an asymptotically regular polygon. These identities yield an explicit finite list of admissible cluster masses and a four-unit mass-gap criterion. We also construct non-simple blow-up solutions on closed Riemannian surfaces in both regimes. Together, these results establish a sharp arithmetic criterion for single-point non-simple CS blow-up on each fixed flat torus when a suitable exterior divisor may be chosen, and show that vortex blow-up need not be simple, answering Li–Liu's question in the negative.

comments
59 pages
fields
Analysis of PDEs, Mathematical Physics
pdf ↗ page arXiv:2609.20042 ↗

We comment on the optimally-transported generalized method of moments (OTGMM) of Schennach and Starck (2026) and give counterexamples to Theorems 2-6. In the small-error analysis, the assumptions do not ensure consistency or asymptotic normality (Theorems 2-3). In the large-error analysis, Theorem 4 states the equivalence between the OTGMM estimator and the GMM estimator with modified moments, and Theorems 5-6 assert consistency and asymptotic normality. Through counterexamples, we show that the equivalence in Theorem 4 fails and a component of the Lagrange multiplier has different limits under the two estimators (Theorems 5-6). We also point out that, under misspecification, the population value selected by OTGMM depends on the metric. We give sufficient conditions and establish results following the original motivation for OTGMM. Our corrections to Theorems 4-6 change their assumptions or interpretation more substantially than those to Theorems 2 and 3. Furthermore, we correct Assumption 16 and the test's null distribution.

comments
Comment on arXiv:2511.05712
fields
Statistics Theory
pdf ↗ page arXiv:2609.20260 ↗

We construct a class of positive cones that make $\mathbb{C}$ into a directed partially ordered ring. The positive cones are defined using integral closures of local rings, associated with a transcendence basis and a chosen real generator. A localization criterion also yields such orders on every transcendental extension of $\mathbb{Q}$. For a fixed coefficient field, we prove that two real generators define the same cone if and only if they differ by an affine map with positive real-algebraic slope and translation algebraic over that field. The nonzero positive elements are closed under inversion.

fields
Rings and Algebras
pdf ↗ page arXiv:2609.20494 ↗

M. E. Zieve conjectured that a candidate datum over the sphere is realizable whenever the gcd of the parts of every branch partition is one and it satisfies the lcm condition. We prove that this conjecture reduces to its three-branch-point case. More precisely, if every Zieve-admissible triple is realizable, then every Zieve-admissible datum with at least four branch points is realizable. As already noted by Petronio, Zieve's conjecture would imply the prime-degree conjecture of Edmonds, Kulkarni, and Stong. Song, Wen, and Zhang proved the latter conjecture in a preprint posted on arXiv in September.

comments
Substantially shortened and reorganized. The proof in Section 3 has been rewritten, and the exposition and references have been updated. The main theorem is unchanged.The number of pages has been reduced from 39 to 23
fields
Differential Geometry
pdf ↗ page arXiv:2609.23663 ↗

Lee-Yang theorems are a powerful tool for studying many-body systems, with applications ranging from analyzing phase transitions to proving the efficiency of certain classical and quantum algorithms. In this work, we prove a Lee-Yang zero-freeness theorem for the partition function of a broad class of interacting fermion models, implying the existence of a provably efficient quantum algorithm for estimating their ground-state energies. This class includes several well-known models such as the attractive Hubbard model, repulsive Hubbard model on bipartite graphs, and the interacting Hofstadter model. Our results also rigorously establish the nonexistence of phase transitions in these models in the presence of a nonzero local external field.

comments
21 pages. v2: extended the class of fermionic Lee-Yang Hamiltonians
fields
Mathematical Physics
pdf ↗ page arXiv:2609.23942 ↗

We provide a generators and relations presentation of pivotal pointed fusion categories, $Vec(G,ω,π)$. Unlike the well-known skeletal model, our presentation is strict and allows multiple isomorphic objects. Our main tool is skein theory, which allows us to apply topological tools to understand the relations of morphisms in the category. This expository paper provides details of skein-theoretic work in the setup that allows multiple isomorphic labels in the generators, which didn't previously exist in the literature.

comments
10 pages. Comments are welcome
fields
Quantum Algebra
pdf ↗ page arXiv:2609.27295 ↗

Many real-world high-dimensional time series exhibit long-memory, but Gaussian graphical model testing in this regime remains understudied. We develop a direct, data-adaptive test statistic for assessing conditional independence in the graph structure of stationary Gaussian time series. We establish a finite-sample, Berry–Esseen type Gaussian approximation bound for the statistic, which applies to both short-memory and long-memory time series. The testing procedure uses a block bootstrap method, on which we provide a finite-sample validity result including in the ultra-high-dimensional scenario, and can be extended to comparing graphical structures in two-sample tests. We also develop a consistency-empowered correction to the statistic and show that such tests attain asymptotic consistency in both size and power. Our proposed method is applied to real-world fMRI data to understand functional connectivities within brain in different periods.

fields
Statistics Theory
pdf ↗ page arXiv:2609.30565 ↗

We study the profinite non-rigidity of arithmetic lattices and its implications for the Kähler property. In the first part, we characterize the absolute Dynkin types that admit non-isomorphic real forms of higher-rank Lie groups containing torsion-free arithmetic lattices with isomorphic profinite completions. In the second part, as a geometric application, we answer a question asked independently by Arapura and Libgober, by showing that the Kählerness of finitely presented, residually finite groups is not determined by its profinite completion.

comments
28 pages, comments welcome
fields
Geometric Topology, Algebraic Geometry, Group Theory
pdf ↗ page arXiv:2609.31523 ↗

The design of satellite constellations for Earth observation requires balancing spatial coverage, revisit time, cost, and operational complexity. This paper considers the problem of designing the orbits of a given number of Low Earth Orbit (LEO) or Very Low Earth Orbit (VLEO) satellites to maximize the spatial and temporal resolution achieved over a prescribed set of ground targets. This kind of problem is inherently nonconvex and possibly combinatorial, making its solution computationally demanding for large constellations and target sets. To address this challenge, we propose a two-stage optimization strategy that separates spatial-coverage design from temporal-resolution optimization, thereby reducing the complexity of the overall problem. Two variants of the method are developed. The first employs continuous decision variables during the spatial-optimization stage, whereas the second discretizes these variables and reformulates the problem as a Quadratic Unconstrained Binary Optimization (QUBO) problem. The latter formulation enables the use of efficient classical QUBO solvers and is directly compatible with quantum-annealing hardware. The proposed framework provides a scalable approach to the design of heterogeneous LEO and VLEO Earth-observation constellations and establishes a pathway for exploiting emerging quantum-optimization technologies in satellite mission design. Preliminary simulation results are presented to demonstrate the effectiveness of the strategy.

comments
17 pages, 1 figure, 1 table
fields
Optimization and Control
pdf ↗ page arXiv:2610.01885 ↗

Beginning with the work of Dinur, Lin, and Vidick (FOCS, 2024), tensor codes with constant product expansion have been foundational to recent advances in quantum locally testable codes (qLTCs) based on cubical complexes. Informally, product expansion says that any low-weight parity check of the tensor code can be written as the sum of few parity checks of each component code in each dimension. Recent work of Kalachev and Panteleev (FOCS, 2025) showed that tensors of random linear codes have constant product expansion. In this work we initiate the study of random linear codes with prescribed automorphism groups. For every sufficiently large prime $k$, we construct pairs of rate $1/2$ MDS codes of length $2k$, over finite fields whose size depends on $k$, that have constant two-sided product expansion and are invariant under the simultaneous cyclic shift of two cycles of length $k$. This shift is the local action in the cube complexes of Rungtanapirom, Stix, and Vdovina (Internat. J. Algebra Comput., 2019). Along the way we give a proof of the existence of two-sided product expanding codes which, unlike the earlier result of [KP25], does not rely on explicit constructions of asymptotically optimal locally testable codes. Our proof relies on several new techniques that we view as interesting in their own right. Instead of working directly over finite fields, we work over the reals and use random Rademacher matrices for the generator matrices of the component codes. From here, we show how to relate product expansion to the operator norm of sparse restrictions of carefully chosen Gram matrices, which we bound using tools from random matrix theory. More generally, we propose a method for studying properties of random linear codes by lifting to random codes over $\mathbb{R}$ and $\mathbb{C}$, which we anticipate will be applicable more broadly to existential problems in coding theory.

comments
100 pages. New results in Part II and fixed typos
fields
Information Theory
pdf ↗ page arXiv:2610.02512 ↗

We denote by $f_k(Γ)$ the largest integer with the property that every subset of a finite abelian group $Γ$ of size at least $|Γ| - f_k(Γ)$ is a $k$-fold sumset. Extending a recent result of Alon and Pham, we prove that $$ f_k(Γ) \leq \widetilde{O} \left(n^{(2k-1)/(4k-3)}\right) $$ holds for all finite abelian groups $Γ$ and integers $k \geq 2$, where $n = |Γ|$. Additionally, we also show that the lower bound $f_k(Γ) \geq \widetildeΩ \left(n^{1/k}\right)$ holds if $Γ$ has no nontrivial element of order dividing $k$. Our upper bound improves a previous result of Balogh, Liu, and Sharifzadeh, and recovers the bound of Alon and Pham in the case $k = 2$. The proof relies on a new upper bound for the independence number of random Cayley sum hypergraphs, which may be of independent interest.

comments
21 pages
fields
Combinatorics, Number Theory
pdf ↗ page arXiv:2610.08629 ↗

We prove that every cluster monomial of type A_n is log-concave and unimodal with respect to its Laurent expansion in a fan initial cluster. We derive a coefficient formula for laminar interval products arising from compatible diagonals in the polygon model. This gives partial affirmative answers to the log-concavity conjecture of Chen-Huang-Sun and the unimodality conjecture of Chen. Our result for cluster A-varieties also complements Shen's study on the supports of products of canonical basis elements for cluster X-varieties of type A_n.

comments
25 pages, 4 figures. Clarified the proofs, improved the exposition, and corrected minor typos. Any comments are welcome!
fields
Representation Theory, Algebraic Geometry, Combinatorics
pdf ↗ page arXiv:2610.09312 ↗

We introduce a secondary index map from relative K-homology to the K-theory of a relative structure algebra. We prove an algebraic Mayer-Vietoris sequence for the relative structure algebra, and prove compatibility of the secondary index map with the Mayer-Vietoris boundary map. As an application, we compute the secondary index of the Dirac operator on an even-dimensional spin manifold with collared metric having uniformly positive scalar curvature on the boundary. This computation leads to a refinement of the delocalized APS-index theorem.

comments
added references
fields
K-Theory and Homology
pdf ↗ page arXiv:2610.09812 ↗

We construct an effective $S_3$ action on the smooth projective model of $y^2=t^{12h-5}+t$ for every $h\geq1$. Its quotient has genus $h$, and the fixed quotient for the three-cycle class consists of two points. We prove that the derived category of these two points admits no $\bbC$-linear exact fully faithful embedding into the equivariant derived category. This gives a counterexample to Conjecture A of Polishchuk and Van den Bergh. More generally, for a positive-genus quotient we compute the maximum number of completely orthogonal exceptional objects as $\Sm_i\lfloor e_i/2\rfloor$, where $e_i$ are the inertia orders. We also give a necessary and sufficient condition for conjugacy-class decompositions with embeddings linear over the quotient, and construct kernels on the natural fibre products. For genus-zero quotients we obtain Euler-form bounds and exact capacities for all-two ramification with at least four branch points and for types $(2,3,3)$, $(2,3,4)$ and $(2,3,5)$.

comments
All comments are welcome!
fields
Algebraic Geometry, Category Theory
pdf ↗ page arXiv:2610.10147 ↗

We study rational approximation to logarithms of positive rational numbers and give a separated-weight determinant argument for $μ(\log r)=2$ whenever $r\in\mathbb Q_{>0}\setminus\{1\}$. The geometric input is an interpolation theorem for logarithmic jets at points with distinct multiplicative coordinates. Its successive weight thresholds are independent of the interpolation centres; the eventual degree threshold may depend on them. A local intersection-length bound and a curve inequality lead to surjectivity through a blow-up and Serre vanishing. For a fixed rational argument, this produces a nonzero rational determinant. Clearing denominators gives an arithmetic lower bound, while an exact row translation and repeated Taylor indices yield a contradictory analytic upper bound. The consequences include nonzero rational linear combinations of such logarithms, rational affine changes, and logarithms of positive numbers with a rational positive integer power. We also give a bounded-fibre interpolation extension, an explicit bound for the final analytic degree, and a counterexample to interpolation under the volume condition alone.

comments
18 pages. Revised manuscript and abstract, with expanded historical context, explanations of the parameter choices, and a discussion of the connection with the argument for pi through the exponential map
fields
Number Theory
pdf ↗ page arXiv:2610.10192 ↗

We show that the characteristic cycle multiplicities of $A$-hypergeometric systems are upper semicontinuous in the parameter for every rational projective weight, thereby proving a conjecture of Schulze and Walther [Duke Math. J. 142 (2008), 465–509]. The proof uses a specialization formula for filtered families satisfying a fixed product support condition, expressing the difference between the special and generic characteristic cycles as the characteristic cycle of the first Tor module. Combined with the Gevrey index theorem, this yields upper semicontinuity of generic Gevrey irregularity and of each of its graded dimensions. We also construct non-Cohen–Macaulay semigroup rings whose Gevrey dimensions are constant in the parameter despite the presence of a genuine irregular slope, showing that the Gevrey jump locus need not coincide with the full exceptional arrangement defined by Ext.

comments
27 pages
fields
Algebraic Geometry, Commutative Algebra
pdf ↗ page arXiv:2610.12330 ↗

We show that any bounded strictly convex domain in $\Bbb C^N$, endowed with the Kobayashi distance, is visible.

comments
to appear in Int. J. Math
fields
Complex Variables
pdf ↗ page arXiv:2610.12502 ↗

We show that finite semi-Markov networks admit explicit combinatorial representations of their Laplace-transformed occupation probabilities and hitting-time distributions as sums over weighted spanning trees and forests of an augmented graph, generalizing the previously developed graph-theoretic framework for Markov networks. These representations yield asymptotic results, including the long-time scaling of hitting-time survival probabilities in the presence of infinite-mean sojourn times. They also extend recently derived combinatorial formulas for hitting-time moments to the semi-Markov setting, where non-Markovian corrections appear in higher moments. For finite-mean sojourn times, the framework recovers the known stationary distribution through a generalization of the Markov Chain Tree Theorem, with transition rates replaced by pseudo-rates. In particular, the mean hitting-time formula coincides with that of a Markov network with the corresponding pseudo-rates, which also yields a short graph-theoretic proof of the previously established constancy of Kemeny's constant for finite-mean semi-Markov networks. Finally, we use the same framework to extend established results on weak ergodicity breaking for time-additive state observables that are nonconstant on the support of the limiting distribution in heavy-tailed semi-Markov networks, allowing for generally heterogeneous, transition-dependent waiting-time distributions.

comments
47 pages, 4 figures
fields
Probability
pdf ↗ page arXiv:2610.13505 ↗

A permutation is $n$-universal if it contains every permutation of length $n$ as a pattern. We prove that a uniformly random permutation of length $32n^2$ is $n$-universal with probability at least $1-\exp(-n/25)$ for all sufficiently large $n$. This removes a $\log\log n$ factor from an earlier result of He and Kwan~\cite{He-Kwan} and resolves an open problem of Alon~\cite{Alon2016}. The proof was found by GPT-6 Pro.

fields
Combinatorics
pdf ↗ page arXiv:2610.13530 ↗

Transmitters in adaptive communication systems may switch their modulation scheme at an unknown time, and a receiver must detect this switch quickly without raising frequent false alarms. Existing modulation classification methods label a finite block of received data and do not provide stopping rules with controlled false-alarm and delay performance. We study the sequential detection of a switch from one phase shift keying (PSK) scheme to another over an additive white Gaussian noise channel. We build cumulative-sum (CUSUM) and moving-sum (MOSUM) detectors from the log-likelihood ratio and its low-SNR approximation. For a switch from quadrature PSK to binary PSK, we derive the moment-generating function and probability density function of the approximate statistic. We use them to relate the CUSUM threshold to the average run length to false alarm and the worst-case detection delay. For a switch between any two PSK orders, we approximate the vector of overlapping MOSUM statistics as multivariate Gaussian and obtain bounds on both measures. Monte Carlo simulations show that our derived expressions for ARL2FA and WADD closely match the empirical results across thresholds, validating the accuracy of the proposed analytical approximations and theoretical bounds.

comments
12 Pages, 5 Figures, Online change-point detection for modulation transitions using CUSUM and MOSUM detectors
fields
Information Theory
pdf ↗ page arXiv:2610.14208 ↗

We develop a computer-assisted approach to Krzyż's conjecture based on validated elimination of the masses in an extremal Herglotz representation. The entropy theorem of Lei and Zhang, together with the Agler–McCarthy reduction, leaves only extremals with fewer than $n$ atoms to exclude. In the critical stratum with $n-1$ atoms, the first-order conditions are linear in the masses. Validated local inversion reduces the regular search to $n$ outer parameters, with a rank-free fallback at singular configurations. This gives a proof of the sixth-coefficient conjecture, including its equality statement; the critical-stratum computation takes 8.2 minutes on an M4 Pro. We also obtain an independent proof for the seventh coefficient, recently established by Xie. The computational exclusions use explicit rounding-error bounds, and the equality cases follow analytically.

comments
18 pages
fields
Complex Variables
pdf ↗ page arXiv:2610.15206 ↗

We analyse how errors in independently approximated adjoint waves propagate into the Hilbert Uniqueness Method (HUM) reconstruction of the minimum-$L^2$-norm interior control of a variable-coefficient wave equation. The central result is an identity: the error of the control assembled exactly from any approximate adjoints splits orthogonally into the distance of the minimum-norm control to their observation span and a moment-inconsistency term generated by their interior wave residuals. From it we derive a posteriori bounds for the returned control, including assembly and solve errors, and a continuous accept-or-refine test for the network directions returned by training; under continuous observability, a fully discrete estimate separates truncation, adjoint and assembly errors. The adjoints are computed here by adaptive Galerkin neural networks, but the bounds apply to any computed adjoints. In one-dimensional benchmarks with known minimum-norm controls, all inputs are enclosed by ball arithmetic. Without truncation the residual bounds exceed the true error by factors of 9 to 115; with sixteen adjoints outside the trial class, a truncated terminal space and all inputs computed from the target data, by about 1.08 to 6.50. In two and three dimensions, including a time-dependent coefficient, the neural controls closely reproduce reference controls of the same truncation, and at equal neuron counts trained hidden features reduce the error more than tenfold relative to random features. Classical solvers are much cheaper in all tested cases, and certification is confined to the one-dimensional benchmarks.

fields
Numerical Analysis, Optimization and Control
pdf ↗ page arXiv:2610.15220 ↗

A model of propagation and accumulation of errors caused by round-off errors during conjugate gradient (CG) iterations is proposed, based on their interpretation as iterations of traversals of the linear system matrix graph. In the matrix graph, vertices correspond to the components of the search direction, residual, and approximate solution vectors, which are computed recursively at CG iterations. It is shown that when the number of CG iterations performed is less than the distance between two vertices in the graph, the round-off errors arising in the computation of one component do not affect the errors in the computation of another and are not correlated. At early iterations, errors propagate locally between neighboring vertices and are locally correlated. As the errors become sufficiently global, loss of orthogonality of residuals and conjugacy of search directions occurs, and stagnation of the iterative process sets in. The results of computational experiments are presented for matrices corresponding to graphs with a prescribed diameter, for Laplacians of differential equations, and for random matrices. These results demonstrate that the diameter and its approximate values, in particular the pseudodiameter, due to their low computational cost for sparse graphs, can be used in mixed-precision algorithms for solving linear systems, namely, for switching precision modes as round-off errors propagate and accumulate.

fields
Numerical Analysis
pdf ↗ page arXiv:2610.15393 ↗

We generalize Moser's stability theorem for minimal hypersurface foliations to arbitrary codimension. Let $\mathbb{T}^N=\mathbb{R}^N/\mathbb{Z}^N$ be the standard torus, and let $P\subset\mathbb{R}^N$ be an $m$-plane satisfying a Diophantine condition relative to $\mathbb{Z}^N$. We prove that, under every sufficiently small smooth $\mathbb{Z}^N$-periodic perturbation of the Euclidean metric, the affine foliation of $\mathbb{T}^N$ induced by $P$ persists as a smoothly conjugate foliation by minimal $m$-dimensional submanifolds. The normalized conjugating displacement depends smoothly and tamely on the metric and is locally unique. Moreover, the perturbed foliation admits a common smooth calibration, hence the lifted leaves are area-minimizing. The proof uses Hamilton's formulation of the Nash-Moser implicit function theorem.

comments
38 pages
fields
Differential Geometry, Analysis of PDEs, Dynamical Systems
pdf ↗ page arXiv:2610.15984 ↗

A set $S$ of positive integers is $r$-large if every $r$-coloring of $\mathbb{N}$ contains arbitrarily long monochromatic arithmetic progressions with common difference in $S$. We prove that every $2$-large set is a set of nil-Bohr recurrence, i.e., it meets every nil-Bohr neighborhood of $0$. This proves a conjecture of the first author and gives a sharp answer to a question of Host, Kra, and Maass. Furthermore, we show that a piecewise syndetic set is $2$-large if and only if it is nil-Bohr recurrent. From this, we show that a small modification of a set constructed by the first author is $2$-large but not $3$-large, thereby disproving the $2$-large is large conjecture of Brown, Graham, and Landman. We also deduce that the family of $2$-large sets is not partition regular, thereby answering another question of Brown, Graham, and Landman. We also show that, for every fixed nil-Bohr neighborhood $B$ of $0$ and every $\varepsilon>0$, restricting the common difference to $B$ increases the extremal number for $k$-term arithmetic progressions by at most a factor of $1+\varepsilon$, provided $k$ is sufficiently large.

comments
20 pages, 1 figure, big update to include resolution of 2-large versus large
fields
Combinatorics, Dynamical Systems, Number Theory
pdf ↗ page arXiv:2610.16605 ↗

We show that the guaranties of the Gram-Schmidt Walk Algorithm of Bansal, Dadush, Garg, and Lovett can be achieved by an Ising model. Given vectors $v_1, \ldots, v_n$ of norm less than 1, we give a closed form for a centrally symmetric distribution on a vector $σ\in \{\pm 1 \}^n$ so that the signed sum $\sum_j σ(j) v_j$ is $1$-subgaussian. We show that the Glauber dynamics for this model mix rapidly provided that the norms of the input vectors are bounded below 1 by an inverse-polynomial. The proofs employ the theory of Lorentzian polynomials, and we use these techniques to show fast mixing bounds for other spin systems.

comments
added results for the Sherrington Kirkpatrick model
fields
Probability
pdf ↗ page arXiv:2610.18639 ↗

Metric spaces of negative type (NT) and strong negative type (SNT) are fundamental in wide applications across computer science and statistics. It is well-known that spheres equipped with geodesic distance are of NT, and subsets of a sphere that contain at most one pair of antipodal points (e.g. an open hemisphere) are of SNT. In this paper, we extend these results to a class of star-shaped manifolds. By constructing a metric $d$ based on layer-wise radial projections, we prove that every star-shaped manifold, equipped with $d$, is of NT. Moreover, we prove that a subset of a star-shaped manifold is of SNT under certain injectivity and directional asymmetry conditions. In particular, for a subset of a single layer, it is of SNT if it contains exactly one point from each preimage of an open hemisphere (a fully directional asymmetry); moreover, it is of SNT if and only if it contains at most one opposite-directional pair. Finally, we prove that under similar conditions, SNT holds for almost and essentially star-shaped manifolds, and for those with incomplete radial coverage.

fields
Metric Geometry
pdf ↗ page arXiv:2610.20023 ↗

We study finite-state approximations of Brownian-driven Markovian backward stochastic differential equations when the same one-step increment is used both for weak propagation and for martingale representation. For a centered isotropic state-independent increment with vanishing third moment, we prove first-order convergence of the Markov value field, its spatial gradient, and the predictable $Z$-field under suitable smoothness assumptions. The corresponding orthogonal martingale has maximal $L^2$ size $O(h^{1/2})$. We identify its quadratic representation defect and prove a joint functional limit theorem for the state process and the normalized orthogonal martingale, which in particular yields a mixed-normal terminal limit with path-dependent conditional variance. On the trace-free quadratic sector, the representation defect has the sharp universal operator lower bound $d/[2(d+2)]$; equality holds exactly for normalized laws that are weighted spherical 4-designs. In dimension $d\geq 2$, vanishing third moment is incompatible with exact scalar one-step predictable representation for any finite-state centered isotropic increment. Consequently, every such exact-representation law has a nonzero cubic direction, and a smooth Fourier terminal condition converts this direction into a sharp global $h^{1/2}$ weak error. For the centrally symmetric regular-simplex rule we also identify the leading quadratic GKW component, compute the exact first-order asymptotic of the accumulated orthogonal-martingale variance, and recover it from the general functional limit. Thus the same third-order tensor links weak accuracy and predictable representation, yielding an accuracy-representation tradeoff for finite-state BSDE approximations.

comments
27 pages; updated a reference and corrected mathematical notation and typography; results unchanged
fields
Probability
pdf ↗ page arXiv:2610.20880 ↗

We develop an operadic theory of anti-splitting and anti-replicating for binary operads. For a binary operad $\opd$, we define its anti-bisuccessor $\ABSu(\opd)$ and anti-trisuccessor $\ATSu(\opd)$. Each generating operation is recovered as the negative sum of two or three split operations, yielding canonical operad morphisms $\opd\to\ABSu(\opd)$ and $\opd\to\ATSu(\opd)$. For binary quadratic operads, these constructions are described by Manin black products: \[ \ABSu(\opd)\cong\antiprelie\bullet\opd, \qquad \ATSu(\opd)\cong\antipostlie\bullet\opd. \] Dually, we also introduce the canonical quadratic anti-duplicator and anti-triplicator, together with their combinatorial counterparts defined by labeled planar binary trees. We show that they coincide precisely under a full-support condition. As applications, we obtain anti-dupli-com and anti-dupli-Lie algebras. We also display an example in which the two constructions differ. Furthermore, we establish the Koszul duality relations \[ \widehat{\ADup}(\opd^!) \cong\ABSu(\opd)^!, \qquad \widehat{\ATrip}(\opd^!) \cong\ATSu(\opd)^!, \] providing an effective description of the canonical quadratic anti-duplicator in terms of the Manin white product. The framework recovers known constructions, including anti-dendriform, anti-pre-Lie and anti-pre-Poisson algebras, and produces further examples and Koszul dual pairs.

comments
comments are welcome
fields
Rings and Algebras, Mathematical Physics, Algebraic Topology
pdf ↗ page arXiv:2610.21365 ↗

Enriques manifolds and Log Enriques varieties have attracted a lot of interest in the last years. In this paper we study Enriques manifolds and Log Enriques varieties that are respectively étale and quasi-étale quotients of Calabi-Yau manifolds. We produce several examples by studying quotients of generalized Fermat manifolds of Calabi-Yau type. These are complete intersections of Fermat type manifolds all of the same degree and were described recently by Hidalgo, Hughes and Leyton-Álvarez. We describe all their quotients by a fixed-point-free automorphism. Most of the quotients are Log Enriques varieties or Calabi-Yau varieties with terminal singularities. In one case we get a family of Enriques manifolds which is a subfamily of a family described previously by Boissière, Camere and the third author, which is one of the two so far known families of Enriques manifolds of Calabi-Yau type. This family is obtained as an étale quotient of order two of a Calabi-Yau manifold complete intersection of quadrics, the other family is the Hilbert scheme of $n$ points on an Enriques surface. In the last part of the paper we show that these two families are different by computing their second Betti numbers, which are equal to one and eleven respectively.

comments
28 pages. Minor changes. Comments are welcome!
fields
Algebraic Geometry
pdf ↗ page arXiv:2610.21752 ↗

We study positive weak singular solutions of the subcritical fractional Hartree equation \[ (-Δ)^s u = \left( \int_{\mathbb{R}^N\setminusΓ} \frac{u^p(y)}{|x-y|^μ}d y \right)u^q \quad\text{in }\mathbb{R}^N\setminusΓ, \] where \(0<s<1\), \(N>2s\), \(0<μ<N\), and \(Γ\subset\{x_1=0\}\) is compact with zero \((s,2)\)-capacity. For solutions with a nonremovable singular point on $Γ$, we prove hyperplane symmetry and strict transverse monotonicity under global Lebesgue integrability conditions. For $N\ge2$, we also prove these conclusions under upper bounds on $p$ and $q$, using only the integrability required for weak solutions. We also establish local regularity and construct solutions with logarithmic singularities along a line segment.

comments
39 pages, comments are welcome
fields
Analysis of PDEs
pdf ↗ page arXiv:2610.24684 ↗

This paper investigates a necessary and sufficient condition for a measurable function on $\mathbb{R}^{\times} \times \mathbb{R}^{d}$ to be the continuous wavelet transform of a complex Radon measure on $\mathbb{R}^{d}$ under suitable regularity conditions on the wavelets. We investigate various integrability properties of the wavelet transform of complex Radon measures. We define a new norm on the collection of such functions and show that it is isometrically isomorphic to the Banach algebra of complex Radon measures under convolution. We show a wavelet inversion formula for measures, in a restricted sense. Under additional hypothesis, we provide a reconstruction formula for the measure of closed rectangles in $\mathbb{R}^d$ using wavelet transforms. We have included supporting examples to substantiate these assumptions.

fields
Functional Analysis
pdf ↗ page arXiv:2610.29293 ↗

We propose a framework to assess the robustness of a given state-feedback controller by evaluating its worst-case finite-horizon cost over a class of norm-bounded perturbations. To compute this cost, we show how to apply the Performance Estimation Problem (PEP) framework, originally developed to analyze optimization algorithms, and reformulate the worst-case search as a rank-constrained Semidefinite Program (SDP). Our results suggest that the perturbations that induce the highest finite-horizon costs can differ substantially from those identified using standard asymptotic metrics. Intuitively, worst-case perturbations are obtained by aligning slowly decaying modes with specific initial conditions to drive transient growth, causing a bifurcation in the worst-case behavior as the horizon length changes. These findings open new possibilities to certify finite-horizon performance of broader system classes, including data-driven models or nonlinear components.

fields
Optimization and Control
pdf ↗ page arXiv:2610.29910 ↗

Every singular subalgebroid, and in particular every singular foliation $\mathscr{G}$, determines a universal Poisson $C^\infty$-scheme $\mathrm{Spec}(\mathcal{F}(\mathscr{G}))$, which is the analogue of the linear Poisson structure on the dual of a Lie algebroid. We show that the underlying space $X_{\mathcal{F}(\mathscr{G})}$ of $\mathrm{Spec}(\mathcal{F}(\mathscr{G}))$ decomposes as a disjoint union of symplectic manifolds. Furthermore, we extend the Helffer-Nourrigat cone to a singular subalgebroid and show that it is an embedded Poisson $C^\infty$-subscheme of $\mathrm{Spec}(\mathcal{F}(\mathscr{G}))$.

comments
27 pages. We have added an AI declaration section and fixed several typos. Comments are welcome
fields
Symplectic Geometry, Mathematical Physics, Algebraic Geometry, Differential Geometry
pdf ↗ page arXiv:2610.30395 ↗

Let $(M^n,g), n\geq 4$ be a close Einstein manifold with positive scalar curvature and $G$ be the Green function of the conformal Laplacian. The ADM mass of the asymptotically flat metric $\hat g:=G^{4/(n-2)} g$ is an important geometric invariant. When either $n=5$ or $n\geq 6$ with Weyl vanishing condition, we prove a mass formula, generalizing Gursky-Malchiodi's formula in dimension $4$. An important step is an extension of Viaclovsky's Green function expansion in higher dimensions. Next we prove a sharp inequality related to the volume radius and the mass for Einstein manifolds. Finally we present two applications. First, we prove a mass gap theorem for Einstein manifolds. Second, for a volume collapsing family of normalized Einstein $n$-manifolds $(M_i,g_i)$ (satisfying the Weyl vanishing condition when $n\geq 6$), we prove that the mass sequence $m(\widehat{g_i})$ blows up uniformly as $i\to\infty$. This in particular applies to Collins-Székelyhidi's family of Sasaki-Einstein metrics $g_i$ on $\mathbb{S}^5$, and to the recent $4$-dimensional examples of T. Ozuch.

comments
27 pages. More references are added, Corollary 2 is added, typos are fixed
fields
Differential Geometry
pdf ↗ page arXiv:2610.31343 ↗

The dominant contemporary paradigm of direct state-led urban environmental regulation entails significant risks, including high administrative burden, frequent conflicts with polluting entities, and economic vulnerabilities for those sources. To address these deficiencies, this paper integrates third-party service provider (TPSP) into China's existing management framework as a beneficial extension beyond conventional government intervention. Utilizing an evolutionary game approach, we compare two distinct institutional settings, a non-integration management mode and an integration management mode, and scrutinize the evolutionary dynamics across both two- and three-dimensional systems. To enhance system robustness, we embed a strategic behavior-based dynamic punishment mechanism that effectively eliminates the perturbing effects of varied penalty rules, thereby steering all stakeholders toward an ideal stable state. This comparative institutional analysis isolates the marginal contribution of third-party intermediation, specifies when outsourcing environmental services outperforms direct government enforcement, and reframes evolutionary game theory as a design device for institutional self-correction. Furthermore, a predefined variable importance analysis is conducted to highlight critical factors warranting attention in practical policy implementation. Extensive simulation experiments, carried out under diverse system-level and parameter-level conditions, validate the theoretical properties of our model. The findings offer concrete managerial insights to inform the formulation of more effective urban environmental supervision regulations.

comments
As the ownership of rights pertaining to this article has not been clarified, the article is hereby withdrawn
fields
Optimization and Control
pdf ↗ page arXiv:2610.32528 ↗

Modeling to Generate Alternatives (MGA) methods explore the near-optimal solutions of energy system optimization models (ESOMs), i.e., alternative planning pathways whose cost lies within a small predefined slack of the optimum but which can differ substantially in technology composition, regional distribution, and environmental impacts. Among MGA methods, ORACLE uniquely guarantees convergence to the near-optimal set within a targeted tolerance by iteratively refining inner and outer polytope approximations. At each iteration, given a trial point, it solves a problem as hard as the ESOM to find the nearest near-optimal solution to it. However, the trial point is obtained by solving a bilevel problem whose solution time grows rapidly with the number of iterations and eventually dominates the computational cost of the method. To eliminate this bottleneck, we present G-ORACLE, which exploits the fact that an optimal solution of the bilevel problem is always attained at a vertex of the outer approximation. G-ORACLE therefore replaces the bilevel problem with incremental vertex enumeration and one convex projection per vertex, while retaining the convergence guarantee of ORACLE. A supporting-hyperplane check further avoids unnecessary recomputation of these projections as the inner approximation grows. We evaluate G-ORACLE on an electricity-only PyPSA-Eur model, exploring the near-optimal space of the installed capacities of six generation and two storage technologies. While ORACLE's bilevel problem exceeds 8 hours at iteration 94, G-ORACLE with the beneath-and-beyond method performs the trial-point search in at most 7 seconds throughout 247 iterations, leaving the ESOM solves as the only significant computational cost.

fields
Optimization and Control
pdf ↗ page arXiv:2610.32759 ↗

The Harary–Hill conjecture states that the crossing number of the complete graph $K_n$ is exactly $\frac{1}{4}\left\lfloor\frac{n}{2}\right\rfloor \left\lfloor\frac{n-1}{2}\right\rfloor \left\lfloor\frac{n-2}{2}\right\rfloor \left\lfloor\frac{n-3}{2}\right\rfloor$. We present a streamlined variant of the proof of this conjecture recently announced by OpenAI (2026). Although the original proof is relatively short, it is quite difficult to follow. We simplify several steps and provide a concise and accessible exposition of the proof. We also combine this new result with known reductions to determine the crossing numbers of several other graph families.

comments
16 pages, updated reference and minor improvements
fields
Combinatorics
pdf ↗ page arXiv:2610.34444 ↗

The univalent Bloch-Landau constant $\mathcal{U}$ is the largest number such that, for every univalent function $f$ on the unit disk, $f(\mathbb{D})$ contains a disk of radius $\mathcal{U}|f'(0)|$. We prove $0.619\le\mathcal{U}<0.6562681$. The previous lower bound, $0.57088586$, came from growth estimates, which control disks centered at $f(0)$, and we show that for Landau-normalized functions such disks cannot give a bound above $0.6$. The main new idea is to move the center of the disk. After Landau's normalization, a compression of the Grunsky matrix constrains every omitted value through the same few coefficients, so a center chosen from these coefficients can be kept away from all omitted values at once. A short analytic proof based on this idea gives $\mathcal{U}\ge0.592$, and a three-row compression, together with pointwise consequences of Landau's normalization, improves the lower bound to $\mathcal{U}\ge0.619$ through an interval certificate. For the upper bound, we construct an explicit domain, bounded by Goodman's first two slit generations, thirty analytic arcs and a circle. We certify $\mathcal{U}<0.6562681$ with a capacity estimate for almost critical arcs, improving the best published bound $0.6563937$.

comments
24 pages, 4 figures. v2: lower bound improved from 0.612242 to 0.619 (computer-assisted, Theorem 1.2(b)). Code, certificates and verifiers: doi:10.5281/zenodo.23981385
fields
Complex Variables
pdf ↗ page arXiv:2610.34580 ↗

We develop and analyze multilevel gradient-regularized Newton methods for unconstrained convex optimization with Lipschitz-continuous Hessians. By computing Newton corrections in lower-dimensional subspaces, the proposed methods aim to reduce computational costs while preserving the global $\mathcal{O}\left(k^{-2}\right)$ convergence rate in function values of the full-space method. To establish this rate for multilevel methods, we introduce a class of row-nested restriction operators that ensures monotonicity of the reduced-gradient norms across levels. Numerical experiments with row-nested operators demonstrate convergence behavior comparable to full-space gradient-regularized Newton, while offering substantial computational savings on large-scale instances.

comments
28 pages, 3 figures
fields
Optimization and Control, Numerical Analysis
pdf ↗ page arXiv:2610.35195 ↗

Lie quandles were defined by T. Fritz as a nonlinear generalization of Lie algebras, to serve as a nonlinear algebras of observables in generalized form of Hamiltonian mechanics. These objects were subsequently generalized in a prequel article to objects referred to as smooth Lie group families of smooth racks by the authors. In this article, we investigate a particular system of nonlinear functional differential equations whose solution in general provides a pathway to the classification of smooth Lie group families of smooth rack structures on any given smooth manifold. We then discuss a particularly practical ansatz for solving these equations, and use it to construct new families of examples, as well as prove a result on the sense in which the Bloch sphere is the unique Lie quandle structure on the sphere.

comments
Some typos were fixed. 29 pages. Comments are welcome
fields
Geometric Topology, Mathematical Physics, Category Theory, Differential Geometry
pdf ↗ page arXiv:2610.35534 ↗

Universally Baire sets are a standard tool in modern set theory, where they are closely connected with forcing and absoluteness. In this paper we study universal Baireness instead as purely a regularity notion for subsets of Polish spaces. Our aim is not to develop the set-theoretic theory, but to place the notion on a systematic descriptive-set-theoretic footing and to clarify the several nonequivalent definitions that occur in the literature. A subset of a Polish space is called \emph{universally Baire} if all its continuous preimages possess the property of Baire. However, it is important to specify what the domain of these continuous functions can be. In fact, a plethora of definitions are used by various authors in the literature, the class of possible domains is sometimes the class of Polish spaces, compact metrizable spaces, all topological spaces, compact topological spaces, etc. We clarify this situation by demonstrating that the above four options are pairwise nonequivalent under the Continuum Hypothesis, we determine all the implications between these four notions that are provable in $ZFC$, and also show that all the notions used in the literature are equivalent to one of these four. Furthermore, we use Miller forcing to show that actually the cases of Polish and compact metrizable spaces are consistently equivalent. We show that numerous definitions using Borel isomorphisms, changes of topologies and Boolean algebras are all equivalent to the case of Polish domains. We present an argument that the notion using Polish domains is the natural dual to the notion of universally measurable sets. Finally, as a byproduct, we prove that the following statement is independent of $ZFC$: if a subset of a Polish space has the property of Baire relative to every nowhere dense Cantor subspace, then the subset has the property of Baire in the whole space.

fields
Logic, General Topology
pdf ↗ page arXiv:2610.36265 ↗

We report a major update to our original software GKMtools and record multiple improvements in the computation of Gromov–Witten invariants of GKM spaces. This includes a refined divisor equation that reduces the complexity from exponential to linear in the number of marked points, three tricks to accelerate the computation of curve class relations in the GKM graph, and an implementation of Mihalcea's algorithm for homogeneous varieties using the equivariant quantum Chevalley formula. As an application, we prove that the small quantum cohomology of the type $E$ generalized Grassmannians $E_7/P_2,E_7/P_4,E_7/P_5,E_8/P_6$ is generically semisimple. These are precisely the remaining exceptional Grassmannians for which generic semisimplicity of small quantum cohomology was still unknown.

comments
20 pages
fields
Algebraic Geometry, Symplectic Geometry
pdf ↗ page arXiv:2610.37273 ↗

In this paper, we prove a zero-density estimate for the symmetric square $L$-functions associated with Hecke eigencuspforms in the weight aspect. The proofs are based on the mollified second moment of these $L$-functions.

comments
27 pages
fields
Number Theory
pdf ↗ page arXiv:2610.37371 ↗

Many problems in nonlinear analysis, game theory, economics, and geometry have existence proofs based on Brouwer's fixed point theorem or the Borsuk–Ulam theorem. We study a common numerical approach: deform a problem with one known solution into the desired problem and follow its zeros. A constructive reduction of fixed points to odd spherical zeros makes the approach available to a broad class of problems, including finite-dimensional Galerkin systems, periodic solutions, Nash and economic equilibria, and geometric mass partitions, subject to the regularity and approximation requirements of each formulation. On the sphere, a linear projection supplies an explicit starting model with one antipodal pair of zeros. We give conditions for tracking the resulting curve, allowing turns in the deformation parameter, and relate computational work to accuracy and conditioning along the path. A family of linear models parametrized by antipodal pairs reaches every regular spherical solution from an open set of starts. This simple parameter space provides a basis for exploring multiple solutions; a finite search for all solutions needs additional coverage information. For mass partitions of finite point sets, each continuation step has cost proportional to the number of points in fixed dimension.

comments
27 pages, 3 figures
fields
Combinatorics, Numerical Analysis
pdf ↗ page arXiv:2610.39092 ↗

We study metric and topological slow entropy for translations on a class of finite-volume $S$-arithmetic homogeneous spaces with real and non-Archimedean factors. We consider translations whose real component is quasi-unipotent and whose non-Archimedean component lies in a compact open subgroup. At polynomial scales, we show that the metric and topological slow entropies coincide and are determined by the chain structure of the real adjoint action. In particular, the non-Archimedean component contributes no additional polynomial slow entropy. Our result extends the slow entropy formula of Kanigowski, Vinhage, and Wei for quasi-unipotent flows on real homogeneous spaces to this $S$-arithmetic setting.

comments
14 pages. Revised version with improvements
fields
Dynamical Systems
pdf ↗ page arXiv:2610.39722 ↗

We develop a state discrimination approach to unassisted quantum communication through a noisy quantum channel. Starting from the operational duality between entanglement recovery and state discrimination, we construct entanglement-generation codes with explicit decoders constructed from Petz recovery maps. The resulting error bounds hold for every number of channel uses and are expressed in terms of Petz-Rényi coherent information. For product sources, they yield exponential error decay at rates below coherent information.

comments
v2: proofs simplified
fields
Information Theory, Mathematical Physics
pdf ↗ page arXiv:2610.40358 ↗

We've heard a lot about AI in math from Fields medalists, AI labs, and established professors. However, the personal stakes are highest for early career researchers. Hoping to provide a platform for graduate students and postdocs to voice their opinions, concerns, and experiences, we wrote and distributed a survey on AI in math. In this article, we share the responses to our survey. All responses are anonymous, and respondents came from almost $60$ institutions and at least $5$ countries. Our survey included both multiple choice questions and more open-ended free response questions. If you are a senior mathematician or university administrator reading these survey results, our goal is for you to understand the concerns of the younger generations and use your authority to take action accordingly. If you are an early career mathematician reading these survey results, we hope you find comfort in the knowledge that your experiences are shared by many others. If you are not a mathematician, we hope you come away from this article with a deeper understanding of the goals, values, and concerns of the community.

comments
11 + 15 pages, 18 figures, fixed small typos
fields
History and Overview
pdf ↗ page arXiv:2610.40584 ↗

We show that all compact hyperkähler manifolds are bigraded strongly formal in the sense of Milivojević and Stelzig, proving that in the context of pluripotential homotopy theory they play the same role as compact Kähler manifolds in rational homotopy theory. Indeed, the main result arises from general algebraic considerations on categories of objects with more differentials, that of $n$-complexes and commutative $n$-differential graded algebras. For them we discuss a standard model category structure, introduce a generalization of the $\partial\bar{\partial}$-Lemma and establish an algebraic formality theorem.

comments
Comments are welcome! Second version: few lines added to the introduction, fixed typos and little notational inconsistencies
fields
Algebraic Topology, Algebraic Geometry, Differential Geometry
pdf ↗ page arXiv:2610.40639 ↗

Many kinds of generalised syntax, including internal categories, T-categories or polynomial monads, arise as monads in bicategories of a certain flavour: maps are some kind of span, and composition is some kind of pullback. We explain the pattern behind these similarities: each bicategory is stitched together from two simpler pieces by a distributive law, where the distributivity is governed by pullback. A key novelty is in the kind of distributive law we consider; where Rosebrugh and Wood considered distributive laws between categories, we instead consider pseudo-distributive laws between double categories. A key construction, drawing on work of Weber, obtains such a distributive law from a so-called co-crossed double category: one whose codomain functor is a fibration, and whose cartesian cells are stable under horizontal composition. We show the horizontal Kleisli double category of a monad T is co-crossed just when T is cartesian, and that the composite double category recovers T-spans; that the double category associated to an identity-on-objects functor is co-crossed just when that functor is a calibration in the sense of Bénabou; and that Street's calibrated bicategories yield co-crossed double categories in a similar manner, recovering the double category of polynomials of Gambino–Kock.

comments
29 pages
fields
Category Theory
pdf ↗ page arXiv:2610.41551 ↗

For a positive rational number $x=p/q$, where $p$ and $q$ are coprime positive integers, let $\operatorname{den}(x)=q$. For a positive integer $n$, let $\operatorname{sd}(n)=\max\{\operatorname{den}(χ_f(G)): |V(G)|=n\}$. Let $D(n)=2^{-n}(n+1)^{(n+1)/2}$. It was proved by Alon, Tuza and Voigt that $\operatorname{sd}(n)\le D(n)$. In this paper, we prove that $\operatorname{sd}(n)\le D(n)n^{-\frac12\log_2 n+\frac14\log_2\log_2 n+O(1)}$ for all sufficiently large $n$. For the lower bound on $\operatorname{sd}(n)$, it was recently proved by Bonamy, Hylasová, Kaiser and Sereni that $\limsup_{n\to\infty}\operatorname{sd}(n)^{1/n}\ge\sqrt{2}$. We prove that $\limsup_{n\to\infty}\operatorname{sd}(n)^{1/n}\ge 3^{1/3}>\sqrt{2}$.

comments
18 pages, 5 figures
fields
Combinatorics
pdf ↗ page arXiv:2610.41771 ↗

We study global compact finite propagation (GCFP) for the wave equation $u_{tt}=\operatorname{div}(A(x)\nabla u)$ on $\mathbb{R}^d$, that is, the property that a disturbance starting in a compact set cannot reach spatial infinity in finite time. For smooth strictly positive matrices $A$ we show that GCFP implies $\int_1^\infty a_*(r)^{-1/2}\,dr=\infty$, where $a_*(r)$ is the minimum over the sphere of radius $r$ of $\langle Aω,ω\rangle$; the proof uses the equivalence of GCFP with completeness of the Riemannian metric $A^{-1}$, of which we prove the necessary direction for the cosine family of the Friedrichs extension. We also prove that the analogous divergence condition with $a^*(r)$, the maximum over the sphere, is sufficient, and that this holds for measurable coefficients $A\in L^1_{\mathrm{loc}}$ with $A\ge I$. Finally, we construct a locally bounded measurable matrix with $a_*=a^*$ for which GCFP holds although the integral converges, so the necessary condition is not stable under the loss of continuity.

comments
10 pages. Misprints corrected
fields
Analysis of PDEs
pdf ↗ page arXiv:2611.00876 ↗

A mutual-visibility set in a graph G is a set S of vertices such that any two vertices of S are connected by a shortest path with no internal vertex in S, and the mutual-visibility chromatic number $χ_μ(G)$ is the minimum number of mutual-visibility sets needed to partition V(G). Babu, Di Stefano and Lakshmanan showed that it is NP-complete to decide whether $χ_μ(G)\le 2$ for graphs of diameter four. We show that the problem remains NP-complete for graphs of diameter d, for every fixed $d\ge2$; diameter two is the smallest diameter for which the problem is non-trivial. More generally, for all fixed $k\ge2$ and $d\ge2$, it is NP-complete to decide whether $χ_μ(G)\le k$ for graphs of diameter d. The proof rests on a reformulation of $χ_μ(G)\le 2$ for diameter-two graphs as the 2-colourability of an associated hypergraph, and on a reduction from 2-colourability of 3-uniform hypergraphs without short cycles.

comments
9 pages, 1 figure. v2: extended to any fixed number k of colours; new title
fields
Combinatorics
pdf ↗ page arXiv:2611.01012 ↗

We prove a mean value theorem on the distribution of smooth numbers in two simultaneous arithmetic progressions, with generalizing a result of Drappeau. Let $\varepsilon>0$ and $(\log x)^C<y<x^{1/C}$ for a sufficiently large $C$. Our result implies that $y$-smooth numbers are well-distributed in the residue classes of a fixed integer $a$ on average over moduli $\leq x^{5/9-\varepsilon}$ and $\leq x^{153/256-\varepsilon}$ for each $|a|\leq x$, which improves Drappeau's range $\leq x^{6/11-\varepsilon}$. As an application, we show that a positive proportion of $y$-smooth numbers $n\leq x$ satisfy $\min(P^+(n-1), P^+(n+1))>x^{407/720}$.

comments
31 pages
fields
Number Theory
pdf ↗ page arXiv:2611.01522 ↗

Let $L/K$ be a degree $p$ extension of $p$-adic fields. We shall provide a complete characterization for the freeness of any fractional ideal $\mathfrak{p}_L^h$, $h\in\mathbb{Z}$, of $L$ as a module over its associated order $\mathfrak{A}_{L/K}(h)$ in the only Hopf-Galois structure on $L/K$, and an explicit description of $\mathfrak{A}_{L/K}(h)$. We show that there is no loss of generality in assuming that $0\leq h\leq p-1$ and that the normal closure of $L/K$ is totally ramified over $K$. If $L/K$ is maximally ramified, then $\mathfrak{A}_{L/K}(h)$ is the maximal $\mathcal{O}_K$-order in $H$ and $\mathfrak{p}_L^h$ is $\mathfrak{A}_{L/K}(h)$-free. The remaining cases correspond to typical extensions. For these, we provide an $\mathcal{O}_K$-basis of $\mathfrak{A}_{L/K}(h)$ and a characterization for the condition that $\mathfrak{p}_L^h$ is $\mathfrak{A}_{L/K}(h)$-free. In the non-exceptional cases, which had been partially studied in the Galois setting, the characterization depends on $h$ and the continued fraction expansion of $\frac{a}{p}$, where $a$ is the residue class mod $p$ of the (fractional) ramification jump $\ell$ of $L/K$. In the remaining exceptional case, where $L/K$ is almost maximally ramified and $a+1\leq h\leq p-1$, it consists in a numerical common minimizer condition.

comments
34 pages
fields
Number Theory
pdf ↗ page arXiv:2611.02301 ↗

We solve the kinetic Kato square root problem for the whole-space Kolmogorov operator $$H_A=\partial_t-X\cdot\nabla_Y-\operatorname{div}_X(A\nabla_X)$$ on \(\mathrm L^2\) with Lebesgue measure, where \(A\) is a bounded measurable complex uniformly accretive matrix depending on all variables. We define its realization through a sesquilinear form and identify the square-root domains of the resulting maximal accretive operator and its adjoint with the intrinsic Sobolev space defined by one velocity derivative and half a derivative along the transport flow. The homogeneous norm equivalences depend only on the dimension and the coefficient bounds. The energy also controls a position derivative of order one third. The proof uses second-order resolvents and localized affine testing. A regularized transport symmetrizer gives hidden coercivity, while comparison with a positive Rockland operator supplies intrinsic smoothing. Commutator and dyadic approximation estimates remove the smoothing before a Carleson argument is applied. We also obtain uniform shifted square-root estimates and bounded velocity, transport and position Riesz transforms.

comments
Revised exposition and expanded declaration of AI assistance
fields
Analysis of PDEs, Functional Analysis
pdf ↗ page arXiv:2611.03104 ↗

We study Hodge–Riemann relations for ordinary, sequential, and weak $m$-positive real $(1,1)$-forms in complex dimension $d$. For ordinary $m$-positive forms, we prove the fully mixed relations in every permitted bidegree without assuming semipositivity, generalizing results by Xiao. As a generalization of ordinary $m$-positivity, sequential $m$-positivity is defined by the hyperbolicity cones of successive directional derivatives of the determinant. We formulate a generalized mixed Hodge–Riemann conjecture for this condition and prove several cases beyond the scalar inequalities supplied by hyperbolicity. We prove strict strong positivity for good prefixes with at most three repeated blocks after the first factor. We also consider weak $m$-positivity which corresponds to the pointwise subsolution condition for the complex $m$-Hessian equation. In this setting, we prove the Hodge–Riemann relations for $p\neq q$ and in several diagonal cases, and show that the associated power map induces a pointwise diffeomorphism onto the cone of positive $(d-1,d-1)$-forms. Using Dinh and Nguyên's $L^2$ globalization method, we pass from the pointwise relations to cohomology classes on compact Kähler manifolds. Finally, we propose conjectures on a numerical criterion for sequential $m$-positive classes and on the solvability of the generalized $m$-Hessian equation.

comments
48 pages
fields
Algebraic Geometry, Differential Geometry
pdf ↗ page arXiv:2611.03124 ↗

Erdős, Rubin and Taylor asked whether $(a,b)$-choosability implies $(am,bm)$-choosability for all positive integers $a,b,m$. Dvořák, Hu and Sereni disproved this for $(a,b,m)=(4,1,2)$ and asked whether a $3$-choosable graph can fail to be $(6,2)$-choosable. Following Jiang and Zhu, Li and Zhu asked whether every $k$-choosable graph is $(kb+1,b)$-choosable for all positive integers $k,b$. Equivalently, Xu and Zhu asked whether $\operatorname{ch}_f^s(G)\le\operatorname{ch}(G)$ for every graph $G$. Here $\operatorname{ch}(G)$ is the choice number, and $\operatorname{ch}_f^s(G)$ is the strong fractional choice number, defined as the infimum of real $r$ such that $G$ is $(a,b)$-choosable for all positive integers $a,b$ with $a/b\ge r$. We prove that, for all integers $t\ge0$ and $b\ge7t+2$, there is a planar graph $G$ with $\operatorname{ch}(G)=3$ that is not $(3b+t,b)$-choosable. This answers all these questions and gives a lower bound of $1/7$ on the supremum of $\operatorname{ch}_f^s(G)-\operatorname{ch}(G)$ over planar graphs of choice number three. For all integers $k\ge4$, $t\ge0$ and $b\ge3t+2$, we construct a graph $G$ with $\operatorname{ch}(G)=k$ that is not $(kb+t,b)$-choosable, using the six-vertex wheel $W_5$ as its core gadget. This gives the corresponding lower bound $1/3$ for each $k\ge4$. The case $t=0$ and $b=2$ recovers the main result of Dvořák, Hu and Sereni. Finally, for every integer $t\ge1$, there are a positive integer $k$ and a $k$-choosable graph that is not $(8k+4t-1,8)$-choosable. Thus the difference between the strong fractional choice number and the choice number is unbounded, even with multiplicity fixed at eight. We further prove $k+Ω(\log k)\le\sup\{\operatorname{ch}_f^s(G):\operatorname{ch}(G)=k\}\le2^{(1+o(1))k}$ as $k$ tends to infinity.

comments
16 pages, 3 figures
fields
Combinatorics
pdf ↗ page arXiv:2611.03236 ↗

Agreement expansion asks whether local assignments to faces of a poset that usually agree on a sampled common face must agree with a single global function on most faces. Agreement expansion is exact (approximate) if local assignments need to exactly (approximately) agree with the global function. We suggest a new proof technique for agreement expansion that enables agreement expansion in settings which were previously beyond reach, such as: exact agreement expansion for Grassmannian posets (before only simplicial complexes support exact agreement expansion); agreement expansion based on assignments to top faces of the a complex (before agreement expansion is generally proven only for assignments to low dimensional faces); agreement expansion from one-sided spectral expansion (before, two sided spectral assumption was needed in general settings). Specifically, for down-closed simplicial and Grassmannian complexes, we prove an exact agreement theorem between any two layers assuming one-sided spectral expansion in links of singletons. As an application, we show that full bases of every rank-$k$ matroid admit exact agreement with soundness $O(k^2/[h(k-h)])$ when tested through $h$-element independent sets. We also obtain exact agreement on Grassmannian posets under one-sided spectral expansion, with stronger soundness under a local distance guarantee. Our key proof ideas center around non-polarity of the distributions of local assignments, probed via a random exposure of elements.

fields
Combinatorics
pdf ↗ page arXiv:2611.03261 ↗

For an infinite Gaussian matrix with independent entries and doubly monotone variances, we identify its algebraic principal ideal by a deterministic shell profile. The exact boundedness condition also implies compactness. A lower singular-value estimate uniform in the mean gives an exact sum law under each fixed compact perturbation. Conditional products and spatial compression yield the corresponding classification for Gaussian evaluations of Hilbert two-forms. For Wiener area of covariance rank at least two, the generator is the covariance tail divided by its index. The area and covariance generate the same ideal precisely under regularity at infinity.

comments
15 pages
fields
Probability
pdf ↗ page arXiv:2611.03495 ↗

In this article we establish new obstructions to the categorification of near-group fusion rings by combining the values of irreducible characters of the Grothendieck ring with the Galois symmetry of the induced unit object $I(1)$ in the Drinfeld center. As a result, we show that a near-group ring of type $G+m$, where $G$ is an abelian group and both $|G|$ and $m$ are odd, can only categorify if $m=|G|$.

comments
11 pages; v2 adds an acknowledgements section, text and mathematical content remain unchanged
fields
Quantum Algebra, Mathematical Physics, Representation Theory
pdf ↗ page arXiv:2611.04232 ↗

The reflective algebra $(R_H(A),K_{\mathrm{ref}}(A))$, introduced by Laugwitz, Walton and Yakimov, is a quasitriangular left $H$-comodule algebra constructed from a quasitriangular Hopf algebra $(H,R)$ and a left $H$-comodule algebra $A$. We study the structure and factorizability of finite-dimensional reflective algebras. When $(H,R)$ is factorizable, we show that $(R_H(A),K_{\mathrm{ref}}(A))$ is isomorphic, as a quasitriangular left $H$-comodule algebra, to $A\otimes H$ equipped with the trivial left $H$-coaction on $A$ and the universal $K$-matrix induced by $R_{21}R$. In this case, we also classify all universal $K$-matrices on $R_H(A)$. Finally, we prove that $(R_H(A),K_{\mathrm{ref}}(A))$ is factorizable if and only if $(H,R)$ is factorizable and $A$ is simple as an algebra.

comments
27 pages
fields
Quantum Algebra
pdf ↗ page arXiv:2611.04326 ↗

Distributed statistical learning has recently attracted significant attention owing to its computational efficiency and scalability in large-scale systems. However, the non-smooth quantile loss poses substantial challenges to the development of efficient decentralized algorithms for quantile regression. Existing methods may either suffer from slow convergence or rely on smoothing-based approximations that require additional stringent assumptions, particularly in high-dimensional settings. To address these issues, we propose a decentralized framework that integrates gradient tracking with iterative hard thresholding for sparse quantile estimation. The proposed method avoids smoothing approximations and relaxes the initialization requirement. Theoretically, we show that the proposed algorithm achieves linear convergence under stated regularity conditions and that the resulting estimator attains the same order of statistical accuracy as its centralized counterpart. We further develop a decentralized information criterion for selecting the working sparsity level and establish support screening guarantees. Numerical experiments and a real-data study illustrate the finite-sample performance of the proposed method.

comments
There is a gap in the proof of the appendix section
fields
Statistics Theory
pdf ↗ page arXiv:2611.04612 ↗

We describe an algorithm and a computer program computing the relative Tate-Shafarevich kernel $Sha^S(K,G)$ of a reductive group $G$ over a global field $K$ with respect to a finite set $S$ of places of $K$ containing no archimedean places. We assume that we know the algebraic fundamental group $Λ=π_1^{alg}(G)$ together with the action on it of the Galois group $Θ=Gal(L/K)$, where $L/K$ is a finite Galois extension with no real places such that $Gal(K^s/L)$ acts trivially on $Λ$. We also assume that we know the decomposition subgroups $Θ_v\subseteq Θ$ for all places $v$ in $S$, and all non-cyclic decomposition subgroups $Θ_v$ for $v$ outside $S$. Under certain conditions, we compute the set $Sha^S(K,G)$ also when $S$ contains archimedean places.

comments
29 pages including 16 pages of a Magma program
fields
Representation Theory, Group Theory, Number Theory
pdf ↗ page arXiv:2611.04773 ↗

We give an elementary proof of a tangent-cone identity which might be known to specialists in geometric measure theory. It concerns Lipschitz mappings $g$ from the closed unit ball $B\subset\mathbb{R}^2$ to $\mathbb{R}^2$. Let $x$ be an interior point of differentiability of $g$ at which $Dg(x)$ has rank one. If the upper one-dimensional Hausdorff density of $g(B)$ at $g(x)$ is at most one, then the tangent cone to $g(B)$ at $g(x)$ coincides with the image of $Dg(x)$. This property is used in the proof of rigidity for the four-gradient problem. We provide a detailed, self-contained proof based on differentiability, connectedness and elementary estimates for one-dimensional Hausdorff measure, without using advanced methods from geometric measure theory.

comments
15 pages, 5 figures
fields
Metric Geometry
pdf ↗ page arXiv:2611.05161 ↗

Using Mac Lane's PROPs, one can introduce multilinear polynomial identities that involve all the operations of a Hopf algebra: the (co)multiplication, the (co)unit and the antipode. We find bases for polynomial identities and calculate their codimensions for the Sweedler algebra, treating it first as an ordinary associative algebra over a field and second as a Hopf algebra.

comments
12 pages; the introduction may overlap with arXiv:2404.15868; a small typo was corrected
fields
Rings and Algebras, Category Theory, Quantum Algebra
pdf ↗ page arXiv:2611.05187 ↗

Using a sharp bound for convolutions of non-negative trigonometric polynomials with the Fejér kernel resembling a pointwise variant of Bernstein's inequality for polynomials, we prove new properties of extremal functions for the Krzyż conjecture. In particular, we resolve a 1993 conjecture of Peretz by showing that if $f$ is extremal for some coefficient, then $|f(0)|\leq 1/e$. Furthermore, we show that if equality holds in this estimate for every extremal function, then the Krzyż conjecture is true.

comments
14 pages. Corrected a typo in the statement of Theorem 3
fields
Complex Variables
pdf ↗ page arXiv:2611.05363 ↗

Bloch's regulator formula for an elliptic curve is often quoted with a generator of the anti-invariant subgroup $H_1(E,\mathbb{Z})^-$ and coefficient $1$ in front of the elliptic dilogarithm of the diamond product. We compare this wording with Brunault's published intersection-product formula. The natural map from the anti-invariant subgroup to the regulator quotient $H_1(E,\mathbb{Z})/H_1(E,\mathbb{Z})^+$ has index $1$ when $E(\mathbb{R})$ has two components and index $2$ when it has one. Thus, up to orientation, the coefficient is $1$ for a quotient generator and $ι(E)\in\{1,2\}$ for a subgroup generator. This is an elementary consequence of the published formula, not a new regulator theorem. We illustrate the distinction on 11.a3 using an externally proved regulator anchor and the Brunault–Bertin special values, and on 20.a4 by high-precision numerical integration. We also correct our earlier quotation of Touafek: his printed identity is $πr=D_E((f)\diamond(g))$, with $r=(2π)^{-1}\int_{γ^-}η$. It is compatible with the subgroup normalization when $Δ<0$; when $Δ>0$ the corresponding coefficient is $1/2$, up to orientation. We do not reassess his isogeny theorem or his final exotic relation here. No new proof of Boyd's conductor-11 split-integral identity is included.

comments
10 pages. v2 corrects our omission of $π$ when quoting Touafek and withdraws the resulting double-normalization criticism; main normalization theorem unchanged. Verification script and log included as ancillary files. No new proof of Boyd's split-integral identity. AI assistance: Kimi (Moonshot AI) and Codex (OpenAI); see declaration
fields
Number Theory
pdf ↗ page arXiv:2611.05420 ↗

A finite set of points in $\mathbb{R}^d$ is isosceles if every three of its points form an isosceles triangle. Erdős asked in 1946 for the largest size $\mathrm{iso}(d)$ of an isosceles set in $\mathbb{R}^d$; the answer was known for $d\le8$. Let $g(d)$ and $M(d)$ be the largest sizes of a two-distance set and of a spherical two-distance set in $\mathbb{R}^d$. We prove that $\mathrm{iso}(d)=\max\{g(d),M(d)+1,M(d-1)+3\}$ for every $d\ge2$, which reduces the problem to these two classical ones. With the recently determined values of $g(d)$ for $9\le d\le22$ we obtain $\mathrm{iso}(d)=\binom{d+1}{2}+1$ for $9\le d\le21$ and $\mathrm{iso}(22)=276$, and we show that the extremal set is unique: for $d\le21$ it is the set of midpoints of the edges of a regular simplex together with its centroid, and for $d=22$ it is the McLaughlin configuration together with its centre. For $10\le d\le22$ every isosceles set with more than $\binom d2+3$ points is a two-distance set or a spherical two-distance set together with its centre. The proofs start from Blokhuis' decomposition of isosceles sets into mutually orthogonal two-distance sets. Counting leaves one configuration open: a two-distance set with $\binom{d+1}{2}$ points in a hyperplane, the largest number allowed by Blokhuis' bound, together with two points that are mirror images of each other. We exclude it by showing that a two-distance set $T$ attaining Blokhuis' bound in $\mathbb{R}^n$ is unisolvent for quadratic polynomials and that, if $T$ lies on two concentric spheres, then apart from two exceptional cases its two layers are spherical $2$-designs whose sizes are determined by the radii. We also show that the squared distances of such a set are in ratio $(k-1):k$ with $(2k-1)^2\le n+1$ if $n\ge9$. Hence there is none for $9\le n\le23$, so that $g(23)\le299$ and $\mathrm{iso}(23)=\max\{g(23),278\}\le299$.

comments
32 pages, 1 table. v2: new title; adds a proof of the formula for iso(d) in all dimensions and an improved upper bound for g(23)
fields
Metric Geometry, Combinatorics
pdf ↗ page arXiv:2611.05421 ↗

We derive maximal mean-square adiabatic estimates for finite-dimensional stochastic Schrödinger equations with Hamiltonian-proportional Wiener noise. For deterministic $C^2$ Hamiltonians whose eigenvalues have constant multiplicities and do not cross on $[0,1]$, we construct a unitary stochastic adiabatic intertwiner driven by the same Wiener process as the exact propagator. A commutator argument, Itô calculus, and Doob's maximal inequality yield an explicit bound for the expected supremum of their squared operator-norm difference on a selected initial eigenspace. Under uniform coefficient bounds, the physical noise scaling $α=σT^{-γ}$, where $T$ is the total evolution time and $σ\geq0$ and $γ\geq0$ are fixed, gives maximal mean-square error $O(T^{-2}+σ^4T^{-4γ}+σ^2T^{-(2γ+1)})$. Consequently, adiabatic convergence holds for every $γ>0$ as $T\to\infty$, while $γ\geq1/2$ recovers the deterministic $O(T^{-1})$ bound for the root-mean-square maximal error. For fixed nonzero physical noise amplitude, the estimate leaves convergence unresolved.

fields
Probability
pdf ↗ page arXiv:2611.06029 ↗

For a cone of reals $x$, the reals of $\mathrm{HOD}^{L[x]}$ are precisely the reals in $Q_3$, or equivalently, the set of reals in $M_1$, the minimal proper class mouse with a Woodin cardinal. We show that if $z$ is any real coding $M_1|δ^{+M_1}$, then $z$ is the base of such a cone (so $M_1^{\#}$ is such a base). We also establish a kind of optimality for this result, by identifying a real $x$ such that $M_1\subseteq L[x]$ but $\mathbb{R}\cap\mathrm{HOD}^{L[x]}\not\subseteq M_1$, although there is a base for the cone in a generic extension of $L[x]$. We prove an analogous result for the reals of $\mathrm{HOD}^{M_{n}(x)}$ for even $n$, as well as an analogous optimality result which holds for all $n$.

fields
Logic
pdf ↗ page arXiv:2611.06032 ↗

We prove that a finite surjective post-critically finite self-map of a variety in characteristic zero remains post-critically finite on every preperiodic integral effective divisor after passing to its periodic image with its return map. Iterating this result proves PCF all the way down. A detailed report on AI usage is contained in the manuscript.

comments
PALOMAR-2026-10-30-000004 v1
fields
Algebraic Geometry, Dynamical Systems, Number Theory
pdf ↗ page arXiv:2611.06106 ↗

In this paper, we are concerned with the construction of global solutions to the Cauchy problem for the one-dimensional non-isentropic compressible Navier-Stokes-Poisson equations, as well as the precise description of their large-time behavior. While results for small initial data are well-established, the case of large initial perturbations remains less satisfactory. Even for constant transport coefficients, only the Nishida-Smoller type solutions have been obtained so far. In this work, we will provide a complete resolution of this problem. The key idea involves revealing favorable hidden identities of the potential force and a new representation for the specific volume, which enables us to derive uniform upper and lower bounds on this quantity. Moreover, this strategy extends to compressible viscous single-ion flow under the Boltzmann relation; notably, this new method for deriving lower and upper bounds on the specific volume also applies to the Navier-Stokes equations. The proof relies on a domain decomposition technique and some new dissipative structures to close the nonlinear energy estimates.

fields
Analysis of PDEs
pdf ↗ page arXiv:2611.06309 ↗

We establish Sobolev–Poincaré inequalities for piecewise $W^{2,p}$ functions in two and three dimensions over families of general polytopic (including shape-regular simplicial and Cartesian) meshes, allowing for degenerate facets. The proposed inequalities feature control by either the broken Hessian or, more naturally for general fourth-order problems, the broken Laplacian. Important tools in their derivation are a Helmholtz decomposition in Banach–Sobolev spaces and Sobolev–Poincaré inequalities for piecewise $W^{1,p}$ functions; we avoid the use of any inverse inequality. We provide estimates with fairly general Lebesgue exponents. Applications to the stability analysis of nonconforming discretizations for second- and fourth-order, linear and nonlinear problems are exhibited.

fields
Numerical Analysis
pdf ↗ page arXiv:2611.06646 ↗

In this paper, we provide an infinite family of counterexamples to Tokareva's conjecture on sums of two bent functions. The conjecture states that every Boolean function in an even number of variables $n$ of algebraic degree at most $n/2$ can be written as the sum of two bent functions. We show that for every $n=2^k\geq 8$, the symmetric function $F_n=1+σ_2+σ_4+σ_6+\cdots+σ_{n/2}$ has degree $n/2$ and cannot be written as a sum of at most two bent functions. Here $σ_j$ is the elementary symmetric Boolean polynomial of degree $j$. Furthermore, we show that the eight-variable member of this family can be represented as the sum of exactly three bent functions. Using the connection between bent-sum decompositions and derivatives of bent functions, we also obtain an infinite family of balanced Boolean functions that cannot be derivatives of bent functions.

comments
18 pages
fields
Combinatorics
pdf ↗ page arXiv:2611.06744 ↗

Get each issue by email →