Differential Geometry
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
Friday 8 May
We prove the lossless unit interval Strichartz theorem on asymptotically conic surfaces, assuming that a large enough neighborhood of its trapped set has negative curvature. We also prove the spectral projection theorem for $6 < q < \infty$ on asymptotically Euclidean surfaces with nonpositive curvature, assuming a large enough neighborhood of its trapped set has negative curvature. If the negatively curved region extends to a sufficiently far-out cutoff, we also obtain the critical and subcritical estimates, without a curvature sign assumption near infinity.
Complex Variables
The idealizer of the semigroup of stable polynomials
It follows from the Garloff-Wagner Theorem that the set of stable polynomials of degree $n$, denoted by $\mathcal{H}_n$, i.e., those whose zeros all lie in the open left complex half-plane, with the Hadamard product $*$, forms an abelian semigroup contained in the abelian group $\mathbb{R}_n^+$ of polynomials of degree $n$ with positive real coefficients. By the idealizer of the set $\mathcal{H}_n$, we refer to the largest subsemigroup of $\mathbb{R}_n^+$ in which $\mathcal{H}_n$ is an ideal. In this paper, we formulate a conjecture characterizing the idealizer of $\mathcal{H}_n$ and prove it for $n \leqslant 5$. In addition, we show that the proposed condition is necessary for any polynomial to belong to the idealizer and establish, in a distinguished special case, a sufficient condition of a similar nature that supports the conjecture.
Read the full abstract
Optimization and Control · Cross-listed from quant-ph
A Unified Local Light-shifts Encoding For Solving Optimization Problems on a Rydberg Annealer
Combinatorial optimization problems play a central role in computer science with many real world applications. A number of relevant problems remain computationally difficult to solve as they lie in the NP-hard complexity class. We present a unified framework for solving such optimization problems represented in the quadratic unconstrained binary optimization (QUBO) formalism, namely two-SAT, XOR-SAT, mixed-two-XOR-SAT, set packing, quadratic assignment, binary clustering, and protein folding, by expanding the domain of applications of PRR, 6(2), 023031. A direct mapping from the QUBO form of these problems onto the Rydberg quantum platform is demonstrated as our first step. This mapping to the Rydberg system depends on distance-dependent long-range interactions and configurable local detuning, thus reducing resource overhead and improving scalability. Following-up on the encoding, the solution is reached by steering the system toward the ground state of the target Hamiltonian using an optimized quantum annealing protocol that controls the time-dependent detuning and Rabi frequency profiles. The framework can handle a variety of problems, each with different complexity. To quantify the complexity of any problem, a generalized hardness parameter is introduced that compares different problems based on the structure of their optimization landscapes. This is a proceedings contribution to the Athens Workshop in Theoretical Physics: 10th Anniversary, held at the National and Kapodistrian University of Athens on December 17-19 2025.
Read the full abstract
Number Theory
Weighted Distributions of Complex Multiplication Orders in Ordinary Isogeny Classes
We develop a global arithmetic framework for studying endomorphism rings inside ordinary elliptic isogeny classes over finite fields. Let p be a prime and let I(t,p) be an ordinary isogeny class over the finite field F_p with Frobenius trace t. The discriminant Delta = t^2 - 4p can be written as Delta = v^2 D_K, where D_K is the fundamental discriminant of an imaginary quadratic field K. In this setting, the possible endomorphism rings are precisely the quadratic orders O_f = Z + f O_K, with f dividing v. Building on Deuring's correspondence, we express the distribution of these orders in terms of weighted class numbers h*(D) = h(D)/w(D), and obtain explicit formulas for global distributions across the entire isogeny class. This approach goes beyond the classical local viewpoint, where the endomorphism ring is constant along each level of an ell-isogeny volcano. In particular, we introduce weighted exact and cumulative distributions of endomorphism rings. These distributions induce canonical laws for the ell-adic valuation of conductors and recover the vertical stratification of ell-volcanoes in an averaged sense. On the global side, by varying the prime p, we relate the existence of curves with a prescribed CM order O_D to splitting conditions in the associated ring class field L_D. Using the Chebotarev density theorem, we obtain the natural density 1/(2h(D)) for primes admitting CM by O_D. This gives a horizontal distribution law complementary to the vertical conductor distribution. These results establish a unified perspective linking Deuring theory, isogeny graph geometry, and class field theory. They also provide a natural framework for quantitative and algorithmic studies of ordinary isogeny classes.
Read the full abstract
Statistics Theory
Statistical Convergence of Spherical First Hitting Diffusion Models
Denoising diffusion models have evolved into a state-of-the-art method for tasks in various fields, such as denoising and generation of images, text generation, or generation of synthetic data for training of other machine learning models. First hitting diffusion models (FHDM) are a particular class of denoising diffusion models with random adaptive generation time tailored to generate data on a known manifold. Building on the conditioning framework of Doob's $h$-transform these models leverage the given information on the target data manifold to demonstrate strong performance across tasks while offering distinct features such as time-homogeneous dynamics of the generating process and a reduced average simulation time. Even though the theoretical investigation of standard forward-backward diffusion models has attracted much attention in the recent past, the statistical convergence properties of FHDMs are not yet understood. In this work, we show that, up to logarithmic factors, FHDMs achieve the minimax optimal convergence rate in total variation for spherically supported Sobolev smooth data distributions. In particular, this is the first statistical optimality result for denoising diffusion modelling with random generation time.
Read the full abstract
Information Theory
Kolmogorov–Nagumo Mean Frameworks for Conditional Entropy
This study focuses on conditional entropy frameworks based on the Kolmogorov–Nagumo (KN) mean. First, $(\eta, \psi)$-KN averaging (EPKNAVG), a KN-mean extension of the $\eta$-averaging (EAVG) framework for $(\eta, F)$-entropies, is introduced and proven to be equivalent to EAVG under suitable concavification conditions. Second, motivated by generalized $g$-vulnerability, a new framework is proposed for generalized $g$-conditional entropies. This framework captures conditional entropies beyond the scope of EAVG-type representations. In particular, it is shown that there exists an $\alpha$ and a joint probability distribution $p_{X, Y}$ such that the Augustin–Csiszár conditional entropy $H_{\alpha}^{\mathrm{C}}(X|Y)$ cannot be represented by any $(\eta,F)$-entropy satisfying EAVG. In contrast, it is represented within the proposed framework. Furthermore, sufficient conditions are derived under which the proposed generalized $g$-conditional entropies satisfy the conditioning reduces entropy property and the data-processing inequality.
Read the full abstract
Algebraic Geometry
The Isomorphism Classes of the Surfaces $x_1^{a_1} + x_2^{a_2} + x_3^{a_3} + 1 = 0$
Let $f = x_1^{a_1} + x_2^{a_2} + x_3^{a_3} + 1 \in \mathbb{C}[x_1,x_2,x_3]$ and let $g = y_1^{b_1} + y_2^{b_2} + y_3^{b_3} + 1 \in \mathbb{C}[y_1,y_2,y_3]$ where $a_1,a_2,a_3,b_1,b_2,b_3 \geq 2$. We prove that the surfaces $V(f) \subset \mathbb{A}^3$ and $V(g) \subset \mathbb{A}^3$ are isomorphic if and only if $(a_1,a_2,a_3) = (b_1,b_2,b_3)$ up to a permutation of the entries.
Read the full abstract
Mathematical Physics · Cross-listed from cond-mat.stat-mech
Hydrodynamics and boundary-induced phase transitions in the $n$-species particle-exchange process
The $n$-species particle-exchange process (PEP($n$)) is an exclusion process in which particles of $n$ different species exchange positions on neighbouring sites with rates chosen such that the invariant measure on the discrete torus is a product measure. We address the large-scale hydrodynamic behaviour of this process which yields a system of $n$ coupled inviscid Burgers equations. This system of conservation laws is shown to admit Riemann invariants for arbitrary $n$ from which explicit solutions of the Riemann problem in terms of shock waves and rarefaction fans are obtained. We also introduce the open PEP($n$), in which particles are exchanged with boundary reservoirs. For a distinguished manifold of boundary rates, we prove that the invariant measure is the same product measure as in the periodic system. The hydrodynamic description in terms of Riemann invariants is used to derive the stationary phase diagram explicitly in terms of microscopic boundary rates. In the generic case, the steady state exhibits $2n+1$ phases, with boundary-induced phase transitions analogous to those of the single-species asymmetric simple exclusion process.
Read the full abstract
Dynamical Systems
Predictive-Switching Control of Stochastic Gene Regulatory Networks: A Contractive PIDE Framework
This paper develops a predictive switching control algorithm for stochastic gene regulatory networks described by a Partial Integro-Differential Equation (PIDE) model, which enables direct shape control of the probability density function. Control inputs are selected from a finite candidate set to minimize a prescribed cost functional. A hybrid framework is proposed for scalability in higher-dimensional systems, using neural networks to approximate the control policy. A central theoretical contribution is a contraction-based analysis of the closed-loop PIDE dynamics. The paper establishes $L^ 1$-contractivity under the proposed control scheme, yielding formal stability guarantees and showing that the evolution of the probability density becomes progressively independent of the initial condition. Moreover, under strictly positive leakage terms, exponential convergence is obtained. The effectiveness and flexibility of the approach, together with the theoretical contractivity results, are illustrated through numerical simulations on three representative examples of increasing dimensionality.
Read the full abstract
Analysis of PDEs
Stationary solutions to the spherically symmetric compressible fluid with capillarity effect
We consider the spherically symmetric Navier–Stokes–Korteweg (NSK) system on the exterior domain $\Omega=\{x\in\mathbb{R}^n~|~|x|>1\}$ with $n\ge2$ when the boundary and far-field data are given. We show that, if the boundary data are sufficiently small, then there exists a unique smooth stationary solution to the spherically symmetric NSK system with impermeable wall, inflow, and outflow boundary conditions. We also establish the decay rate of the stationary solutions. Precisely, the stationary solution for the impermeable wall problem exponentially decays to the far-field states, while that of the inflow/outflow problem algebraically decays. Finally, we investigate the asymptotic convergences of the stationary solution for the impermeable wall problem as the capillarity coefficient vanishes. Numerical results validate that our theoretical convergence rate of the stationary solution is optimal.
Read the full abstract
Group Theory
Solvability and Rigidity for Topological Skew Braces
We study compact and locally compact topological analogues of the Byott–Vendramin solvability problem for finite skew braces, asking whether solvability of the additive group forces solvability of the multiplicative group. Our main theorem proves an affirmative result in the connected locally compact Hausdorff setting: if \(B=(B,\cdot,\circ)\) is a connected locally compact Hausdorff topological skew brace and the additive group \((B,\cdot)\) is solvable, then the multiplicative group \((B,\circ)\) is solvable. The proof proceeds by reducing the additive group to a solvable Lie quotient and then applying an affine-action theorem: a connected Lie group acting transitively and affinely on a connected solvable Lie group, with solvable stabilizer identity component, is itself solvable. We further show that the Hausdorff, local compactness, and connectedness hypotheses are essential by constructing counterexamples when each is omitted. In the compact connected Hausdorff case with abelian additive group, we obtain a stronger rigidity phenomenon: the two group laws coincide.
Read the full abstract
Analysis of PDEs
Uniqueness for an inverse coefficient problem of a weakly coupled parabolic system
This paper considers the weakly coupled parabolic system $\partial_t u-\partial^2_xu +P(x)u=0$ with the homogeneous Neumann boundary condition, where \(P(x)\) is a \(2\times2\) symmetric real-valued function matrix. Under the assumption that the initial value \(a(x)\) is a generating element (i.e., it has a nonzero inner product with every eigenfunction), we prove that the coefficient matrix $ P(x)$ is uniquely determined by the boundary observation $u(0, t)$, $u(1, t)$, $0 < t < T$. The proof relies on the eigenfunction expansion of the solution to the initial-boundary value problem and an extension of the Gel'fand-Levitan theory to the parabolic system.
Read the full abstract
Complex Variables
The Pseudo-Analytic Mass of a Beltrami-Vekua Equation
Every smooth first-order real planar elliptic system admits a universal complex form $w_{\bar z} - \mu w_z + \mathcal{A} w + \mathcal{B} \bar w = \mathcal{F}$, which we call the Beltrami-Vekua equation: the data $(\mu, \mathcal{A}, \mathcal{B}, \mathcal{F})$ are produced from the original system by algebraic operations and differentiations, with no auxiliary PDE. On this space we study the joint action of multiplicative gauges $w \mapsto \phi w$ and orientation-preserving diffeomorphisms. Our main result is that the 2-form $\Theta = |\mathcal{B}|^2 / (1 - |\mu|^2) \, dx \, dy$ is gauge-invariant and pulls back covariantly under diffeomorphisms; its form is forced, with $|\mathcal{B}|^2$ the unique $\mathcal{B}$-quadratic combination invariant under $\mathcal{B} \mapsto \mathcal{B}\phi/\bar\phi$ and $1 - |\mu|^2$ the conformal distortion factor from the diffeomorphism law for $\mu$. The total mass $\mathcal{M}(D) = \int_\Omega \Theta$, the pseudo-analytic mass, vanishes precisely on the analytic class $\mathcal{B} \equiv 0$ and separates a continuous family of pairwise inequivalent pseudo-analytic equations on the disk. As a by-product, Vekua's two-stage reduction - uniformization then gauge elimination - requires only one variable-coefficient PDE solve: the Beltrami diffeomorphism supplies the integrating factor for a flat $\bar\partial$-equation.
Read the full abstract
Information Theory
A Syndrome–Space Approach to Proximity Gaps and Correlated Agreement for Random Linear Codes and Random Reed–Solomon Codes
Proximity gaps and correlated agreement have become central tools in the analysis of interactive oracle proofs of proximity (IOPPs) and code-based SNARKs. Informally, a proximity-gap statement says that for a structured set of words – such as an affine space, or a curve – either all points are close to the code, or most are far from it. Such statements are essential in sampling-based proof systems, where a verifier queries only a few random locations on a structured object but must still obtain a global soundness guarantee. In Reed–Solomon-based proof systems, one would ideally like the proximity parameter to approach the information-theoretic limit $1-R$, since this is the largest possible radius for a rate-$R$ code and directly affects protocol efficiency. We establish a direct approach to proximity gaps and correlated agreement for random linear codes in the random parity-check-matrix model, without relying on list decoding of the proof. Our approach is based on a syndrome-space reformulation together with a witness-based reduction argument. It is conceptually different from the existing decoding-driven route for random linear codes, and it also leads to sharper parameters, including the optimal-up-to-$\varepsilon$ large-alphabet radius bound $\rho<1-R-\varepsilon$ for $q=\Theta(n)$, as well as near-capacity bounds over constant alphabets with improved alphabet-size requirements. We apply the same syndrome-space reductions to random Reed–Solomon codes. This yields correlated agreement for random Reed–Solomon codes over affine spaces and polynomial curves up to radius $\rho\le 1-R-\varepsilon$, with field size $q\ge n\cdot 2^{O(\varepsilon^{-3})}$ for affine spaces and $q\ge n\cdot 2^{O_\ell(\varepsilon^{-3})}$ for degree-$\ell$ curves.
Read the full abstract
Probability
Nonnegativity of the second largest eigenvalue of $4 \times 4$ tridiagonal stochastic matrices
The spectral study of nonnegative and more specifically stochastic matrices is an important topic in matrix theory. In this paper, we prove a conjecture, formulated by Ran and Teng, which states that the second largest eigenvalue of an irreducible $4\times4$ tridiagonal stochastic matrix is nonnegative. We establish this conjecture and extend the result to arbitrary $4\times4$ tridiagonal stochastic matrices, including both irreducible and reducible cases.
Read the full abstract
Combinatorics
A Combinatorial Framework for the Pons-Batle Identity: Young Tableaux, Lattice Paths, and Limit Laws
Tree-child networks are an important class of phylogenetic network used to model reticulate evolutionary processes. These networks have attracted increasing attention from researchers with interests in both combinatorics and algorithms. A fundamental open problem posed by Pons and Batle asks whether the number $TC_{n,k}$ of bicombining tree-child networks with $n$ leaves and $k$ reticulation nodes equals the number of certain constrained words, now called Pons-Batle words. In this paper, we confirm the conjecture for tree-child networks with a bounded number of reticulation nodes. Our approach is combinatorial and analytic. We introduce families of Young tableaux with walls and holes and construct explicit bijections with Pons-Batle words, yielding a direct combinatorial explanation of the identities. These tableaux encode structural features of the underlying networks, including the placement of reticulation nodes. By projecting them to decorated Dyck paths, we obtain algebraic generating functions with differential operators encoding step weights, leading to explicit recurrence relations and closed-form formulas for $TC_{n,k}$. Beyond finite verification for moderate $k$, the framework reveals an underlying probabilistic structure. For $k=1$, natural structural parameters, such as the position and value of distinguished cells, converge, after rescaling, to $\mathrm{Beta}(2,1)$, $\mathrm{Beta}(1,2)$, and Uniform (i.e., $\mathrm{Beta}(1,1)$) distributions. These limit laws arise from a coalescence of singularities at the dominant square-root singularity, producing a non-analytic transition in the local expansion. Overall, our results provide both combinatorial insight and a unified analytic perspective on the asymptotic behavior of tree-child networks, showing how algebraic generating functions with interacting singularities systematically produce Beta limit laws.
Read the full abstract
Complex Variables
Avoidance Criteria for Normal Holomorphic Curves on Complex Projective Space
We establish an avoidance criterion for families of holomorphic curves from the unit disk in complex plane to the complex projective space that omit sufficiently many moving hypersurfaces in pointwise general position. Furthermore, we study families of holomorphic curves that share hyperplanes and derive analogous normality conditions in this context.
Read the full abstract
Number Theory
A Deterministic Cryptographic Prime Generation Chain over Monogenic Cubic Number Fields and their Generalizations
Generating primes is a fundamental problem in modern cryptography. Deterministic primality tests work well for special integers such as Mersenne or Proth primes, but these forms are quite restrictive. In this paper, we give a direct method to construct new primes from known ones. Starting with a seed prime $q \equiv 1 \pmod{3}$, we construct an integer $N \equiv 1 \pmod{3}$ satisfying $(2N + 1)^2 \equiv -3 \pmod{q}$. We then prove that $N$ is prime using the structure of monogenic pure cubic fields $K = \mathbb{Q}(\sqrt[3]{d})$. The resulting test requires only a single modular exponentiation and runs in $\tilde{\mathcal{O}}(\log^2 N)$ time. Finally, we show how this construction extends to pure number fields of arbitrary prime degree.
Read the full abstract
Classical Analysis and ODEs
On ratios of theta functions
Motivated by the average partition function of c free bosons $($Afhkami-Jeddi et al. \cite{Afhk2021}$)$ and the average of the genus 1 partition function over the Narain moduli space $($Maloney-Witten \cite{Witten2020}$)$, we investigate ratios of theta functions. In this paper, we completely classify the minimizers (or maximizers) for ratios of theta and Epstein zeta functions. We find that the hexagonal lattice plays a pivotal role there. These results have direct applications in conformal and Liouville field theory via partition functions. Additionally, they yield the minima of differences of theta and Epstein zeta functions, which have implications for the mathematics of crystallization and interacting particle theory (\cite{Bet2016,Bet2019AMP}).
Read the full abstract
Numerical Analysis
On structure-preserving and pointwise conservative continuous DG schemes for hyperbolic systems
We present a new class of structure-preserving semi-discrete continuous-discontinuous Galerkin (CG-DG) finite element schemes for linear and nonlinear hyperbolic systems of partial differential equations on unstructured simplex meshes that automatically satisfy the following properties: i) the new schemes are not only cellwise conservative, but also locally pointwise conservative everywhere, hence they satisfy the integral form of the conservation law on arbitrary control volumes that do not have to coincide with the mesh at all; ii) the new methods naturally satisfy the two basic vector calculus identities $\nabla \cdot \nabla \times \mathbf{A}$ and $\nabla \times \nabla Z$ exactly pointwise locally and globally everywhere on the discrete level; iii) for linear symmetric hyperbolic systems the schemes are naturally energy conservative for the square energy, i.e. nonlinearly stable in the $L^2$ norm. The key ingredient of the new CG-DG schemes is the use of two different but compatible approximation spaces: the classical DG space $\mathcal{U}_h^N$ of discontinuous piecewise polynomials of degree up to $N$ and a classical finite element space $\mathcal{W}_h^{N+1}$ of globally continuous piecewise polynomials of degree $N+1$. In the new CG-DG schemes, the discrete solution $\mathbf{u}_h$ is sought in $\mathcal{U}_h^N$, while a suitable discrete flux field $\tilde{\mathbf{f}}_h$ is computed in $\mathcal{W}_h^{N+1}$. For $N=0$ our new schemes are directly related to cell-centered finite volume schemes with suitable vertex-based fluxes. All claimed properties of the schemes are first mathematically proven and are then also verified via suitable numerical tests. We show applications of our approach to three linear and nonlinear hyperbolic systems.
Read the full abstract
Algebraic Topology
Combinatorial Models for Linear Homotopy Theories
For a field $k$ of characteristic $0$, we compare $k$-linear chain complexes, semisimplicial vector spaces, augmented semisimplicial vector spaces, semicubical vector spaces, and arboreal vector spaces through small differential categorical algebras. We prove that semisimplicial modules and augmented semisimplicial modules are equivalent to appropriate chain-complex homotopy theories, both at the Gabriel–Zisman localization and the Quillen model-categorical level. The semicubical sign embedding gives a natural comparison from semicubical modules to augmented semisimplicial modules and induces a Quillen adjunction, but not a Quillen equivalence on the full semicubical category since there is an obstruction in augmented homology at degree $-1$.
Read the full abstract
Probability
On the Besov-Orlicz path regularity of some Gaussian processes
In this paper, we rely on the additive decomposition in law satisfied by a class of stochastic processes, combined with the well-known regulariy properties of fractional Brownian motion, to establish Besov-Orlicz regularity of their sample paths. This provides a unified and direct proof for a broad class of processes, including bifractional Brownian motion with parameters $H\in (0, 1]$, $ K\in (0, 2)$ such that $HK \in (0, 1)$, subfractional Brownian motion with Hurst parameter $H\in (0, 1)$, and certain class of self-similar processes. %associated with the stochastic heat equation.
Read the full abstract
Functional Analysis
Schur bounded patterns, submajorisation and operator Lipschitz functions
A Schur bounded pattern is a subset $S\subset \mathbb{N}^2$ such that Schur multiplication by every bounded function on $\mathbb{N}^2$ supported on $S$ defines a bounded linear operator in the norm of $\mathcal{B}(\ell_2(\mathbb{N})).$ Schur bounded patterns were characterised by Davidson-Donsig as being unions of row-bounded and column-bounded sets. We study the analogous question for sets $S$ such that element-wise multiplication by every bounded function on $S$ is bounded in ideals of compact operators that are not closed under submajorisation, in particular the Schatten ideals $\mathcal{L}_p$ with $0<p<1$ and the weak Schatten ideal $\mathcal{L}_{1,\infty}.$ Conversely we characterise the ideals that are not closed under submajorisation by their Schur bounded patterns. This has implications for the functions which are Lipschitz in the norm of ideals that are not closed under submajorisation. In particular such functions must be differentiable and have derivative that is asymptotically constant at infinity.
Read the full abstract
Functional Analysis
On hypercyclic and (common) upper frequently hypercyclic subspaces
Let $B$ be a unilateral weighted backward shift on $\ell_p$, $1 \leq p < \infty$, that admits a $\mathscr{U}$-frequently hypercyclic subspace. We prove that $B$ admits such a subspace free of frequently hypercyclic vectors. The proof technique we develop also allows us to prove that $B$ admits a hypercyclic subspace free of $\mathscr{U}$-frequently hypercyclic vectors, and to solve a question posed by Bès and Menet in 2015 on the existence of common $\mathscr{U}$-frequently hypercyclic subspaces.
Read the full abstract
Representation Theory
Limits and colimits in silting theory with applications to the wall and chamber structure of an algebra
In this paper we consider a family of nested t-structures given by silting objects and construct a silting object corresponding to the intersection of aisles of these t-structures as a homotopy colimit. The dual construction for the cosilting case is given as a homotopy limit. The results are applied to construct two-term large silting objects corresponding to the numerical torsion pairs and the limiting walls in the wall and chamber structure of the real Grothendieck group of a finite dimensional algebra. In particular, in case the algebra is tame we can describe any numerical torsion pair in this way by combining our results with results of Plamondon and Yurikusa.
Read the full abstract
Analysis of PDEs
Stability of the ball in isoperimetric inequalities between two fractional perimeters
We consider the isoperimetric inequality involving the $s$-perimeter and the $t$-perimeter with $0<s<t<1$, and show that the ball is a local minimizer of the (scale-invariant) isoperimetric ratio $\mathcal{F}(E):=P_t(E)^{\frac{1}{n-t}}/ P_s(E)^{\frac{1}{n-s}}$ among sets $E$ that are nearly spherical. To this end, we rewrite $\mathcal{F}$ as a functional of $u$, where $u$ is a scalar function on the unit sphere in $\mathbb{R}^n$ that parametrizes the boundary of $E$, and prove a quantitative stability result for $\mathcal{F}$ around $u=0$ with respect to a suitable Sobolev norm. This parallels known results where the $s$-perimeter is replaced by the volume.
Read the full abstract
Combinatorics
Brik's sequence: a strange recursion
We study the properties of the sequence of words $(B_i)$, where $B_1 = 101$ and $B_{i+1} = B_i C_i$ for $i \geq 1$, where $C_i$ is $B_i$ with the first $i$ symbols removed, and the infinite binary sequence ${\bf b} = 10101101011011101 \cdots$ of which all the $B_i$ are prefixes. We show that $\bf b$ is recurrent, but not uniformly recurrent; it has exponential factor complexity; it is not morphic; and the density of $1$'s exists and is transcendental.
Read the full abstract
Analysis of PDEs
Nonlinear stability threshold for 3D compressible Couette flow
We establish the nonlinear stability threshold $O(\nu^{3/2})$ for the three-dimensional Couette flow governed by the compressible Navier–Stokes equations. While stability thresholds are well understood in two dimensions for both compressible and incompressible flows, and in three dimensions for incompressible flows, the three-dimensional compressible case remains open due to additional structural features, strong mode interactions, and wave coupling. The proof is based on a refined frequency-space approach. For zero modes, we improve upon two-dimensional methods by clearly separating and precisely estimating the main contributions from diffusion waves, acoustic waves, and the lift-up mechanism, leading to a systematic way to handle their nonlinear coupling. For the non-zero modes, we introduce new multiplier estimates and a decomposition based on the structure of the compressible system, which allows us to track the interaction between dissipation and acoustic effects.
Read the full abstract
Analysis of PDEs · Cross-listed from nlin.SI
The Korteweg-de Vries limit for the global dynamics of the Toda lattice
It has been observed that the dynamics of the Toda lattice can be well described by solutions of the Korteweg-de Vries (KdV) equation in the continuum limit. We show that, under the KdV scaling and a suitable translation, the solution of the Toda lattice with H^1 initial data converges to that of the KdV equation globally in time. Our proof relies on tools from harmonic analysis and also on the construction and the conservation of mass and energy of the Toda lattice, the latter of which are derived from the completely integrable structure of the Toda lattice. As a consequence, we obtain long-wave KdV limits for the Toda lattice.
Read the full abstract
Optimization and Control · Cross-listed from cs.LG
SGD for Variational Inference: Tackling Unbounded Variance via Preconditioning and Dynamic Batching
Black-Box Variational Inference (BBVI) typically relies on Stochastic Gradient Descent (SGD) to optimize the Evidence Lower Bound (ELBO). However, the stochastic gradients in BBVI inherently exhibit unbounded variance, violating standard assumptions and instead satisfying the weaker Blum-Gladyshev (BG) condition, where variance grows quadratically with distance from the optimum. In this paper, we bridge the gap between stochastic optimization theory and the practical instances of BBVI. Focusing on the broad elliptic location-scale family of parameterized distributions, we offer two main contributions. First, we prove the existence of an ELBO solution, a foundational property usually assumed a priori in the literature. Second, we establish comprehensive convergence guarantees spanning finite-time and asymptotic regimes for Minibatch Projected SGD (PSGD) equipped with dynamic batching and preconditioning under the BG condition. Our theoretical framework demonstrates that dynamic batching combined with preconditioning systematically enables rigorous guarantees even in complex settings. We illustrate our theoretical findings with numerical results, highlighting the efficacy of our approach for modern inference tasks.