Peer Revenue Publish or profit

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

Mathematics

math Watch Mathematics (sign in to use)

Wednesday 1 April

Classical Analysis and ODEs

Maximal functions with twisted structures, distribution inequality and applications

Motivated by the geometric reduction of Cauchy–Szegő projections on quadratic surfaces of higher codimension (Nagel–Ricci–Stein, 2001) and recent developments on the real-variable theory adapted to twisted multiparameter structures (arXiv:2603.26119), we establish the Fefferman–Stein type distribution inequality relating the twisted area function and the twisted non-tangential maximal function over $\mathbb{R}^{2m}$. By deploying a recursive integration-by-parts argument involving the twisted gradient and Laplacian, and constructing smooth, compactly supported weight functions to absorb cross-derivative errors, we obtain the required estimate. As an application, we prove the uniform $L^1$ boundedness of the twisted maximal function on the twisted atoms and complete the maximal function characterization of the twisted Hardy space.
Read the full abstract

All authors Ji Li, Chong-Wei Liang, Chaojie Wen, Qingyan Wu

Numerical Analysis

A Residual Minimization approach for Nonlinear Partial Differential Equations set in Banach spaces

In this work, we propose and analyze a residual-minimization strategy for the numerical solution of nonlinear PDEs posed in Banach spaces. Given a finite-dimensional trial space and a suitably enriched discrete test space (of higher dimension than the trial space), we approximate the solution by minimizing the variational residual in a discrete dual norm. This minimization is equivalent to a nonlinear saddle-point formulation for the discrete solution in the trial space together with a residual representative in the test space. The latter provides a natural a posteriori error estimator, enabling automatic mesh adaptivity. To solve the resulting nonlinear saddle-point problem, we propose a Newton iteration whose linearized saddle-point system is symmetric, thereby guaranteeing solvability at each step. We take the $p$-Laplacian as a model problem and support the theoretical developments with representative numerical experiments, using standard $H^1$-conforming piecewise linear functions for the trial space, and lowest-order Crouzeix–Raviart functions for the test space.
Read the full abstract

All authors Ignacio Muga, Jorge Perera, Sergio Rojas, Ricardo Ruiz-Baier

Optimization and Control

Extended State Observer for Localized Fault Awareness in RF Accelerating Structures

An observer framework is presented for robust regulation of RF cavity fields and localized identification of disturbances in RF systems. A standard cavity field observer is augmented with additional states to estimate the evolution of cavity detuning and phase drifts induced by the drive and receiver chains. Monte Carlo simulations are performed to assess the performance of the proposed estimator under realistic conditions for the intended high-power linear accelerator operation. Results showcase precise cavity field regulation and the reliability with which the observer assigns deviations to the correct subsystem. The resulting diagnostic capability provides a foundation for improved fault detection, faster troubleshooting during accelerator operation, and more informed maintenance of RF systems in large accelerator facilities.
Read the full abstract

All authors Luke S. Baker, Sungil Kwon, Kwame Jyamfi, Quinten Cole, Isaac Roybal, AJ Garcia, Phil Torrez, Lawrence Castellano

Statistics Theory

E-Values, Bayes Risk, Dual Role of Markov's Inequality

Two approaches to hypothesis testing, e-value testing and Bayes risk minimisation, both invoke Markov's inequality to control error probabilities. They differ in which distribution certifies the unit-moment condition: the null for Type I error, the alternative for Type II error. The likelihood ratio is not intrinsically an e-value; it acquires that status only relative to the experiment under which its expectation is certified. This note makes the resulting role-reversal symmetry explicit, traces its asymptotic sharpening through the information-theoretic arguments of Barron and Clarke (1994), and situates the duality within the typed evidence calculus of Polson, Sokolov, and Zantedeschi (2026).
Read the full abstract

Analysis of PDEs

Existence of Complementary and Variational Weak Solutions to Obstacle Problems for a Quasilinear Wave Equation

We prove the existence of weak solutions for the one obstacle problem associated with a class of quasilinear wave equations in one space dimension, extending previous results obtained in the linear case, and we also address the two obstacles problem. In contrast with the linear setting, for both strictly quasilinear cases we obtain continuous solutions in a weak complementary sense, which moreover satisfy a weak entropy condition in the free region where the string is not in contact with the obstacles. We further show that, in both the one and two obstacle cases, these solutions are variational solutions in a hyperbolic sense without the viscosity term.
Read the full abstract

Algebraic Topology

Maximal compatibility of disklike $G$-transfer systems

Transfer systems are a combinatorial model for $N_{\infty}$-operads, which encode commutative structures in equivariant homotopy theory. Blumberg–Hill and Chan gave criteria for when two transfer systems are a compatible pair, meaning they encode the additive transfers and multiplicative norms of a ring-type structure. In this paper, given a transfer system encoding an additive structure, we give explicit formulae for the maximal transfer system it is compatible with. Our formulae simplify for disklike transfer systems, which typically encode additive structures. Further, we prove that (maximal) compatibility is functorial with respect to the inflation map induced by a quotient of groups, letting us compute maximal compatible transfer systems as inflations of connected transfer systems.
Read the full abstract

All authors David DeMark, Michael A. Hill, Yigal Kamel, Nelson Niu, Kurt Stoeckl, Danika Van Niel, Guoqi Yan

Numerical Analysis

MVNN: A Measure-Valued Neural Network for Learning McKean-Vlasov Dynamics from Particle Data

Collective behaviors that emerge from interactions are fundamental to numerous biological systems. To learn such interacting forces from observations, we introduce a measure-valued neural network that infers measure-dependent interaction (drift) terms directly from particle-trajectory observations. The proposed architecture generalizes standard neural networks to operate on probability measures by learning cylindrical features, using an embedding network that produces scalable distribution-to-vector representations. On the theory side, we establish well-posedness of the resulting dynamics and prove propagation-of-chaos for the associated interacting-particle system. We further show universal approximation and quantitative approximation rates under a low-dimensional measure-dependence assumption. Numerical experiments on first and second order systems, including deterministic and stochastic Motsch-Tadmor dynamics, two-dimensional attraction-repulsion aggregation, Cucker-Smale dynamics, and a hierarchical multi-group system, demonstrate accurate prediction and strong out-of-distribution generalization.
Read the full abstract

Tuesday 31 March

Optimization and Control

The Regional Boundary Reconstruction Problem of the Initial State for Fractional Semilinear Systems

Observability is a fundamental concept in control theory. Its primary purpose is to look into whether it is possible to reconstruct the system's initial state using only the information from its outputs. This paper focuses on the regional reconstruction problem of the initial state for a semilinear time-fractional system, which refers to the possibility of recovering the value of the initial state on a desired boundary sub-region instead of the whole evolution domain or its boundary. To achieve this objective, we use the analytical method, where we suppose that the system's dynamic generates an analytic semigroup. First, by establishing an internal sub-region, we establish a connection between the concepts of regional observability and regional boundary observability; we will later explain how to define the internal sub-region. Then, by imposing suitable assumptions on the analytic semigroup and the system's non-linearity, we give the main theorems of this research from which we deduce a sequence that converges to the unknown initial state on the desired boundary sub-region. Moreover, we present an algorithm that produces some numerical simulations which align closely with our theoretical findings.
Read the full abstract

Numerical Analysis · Cross-listed from eess.SP

On the Instantaneous Phase and Frequency Estimation of a Non-Stationary Multicomponent Signal. The JADE Algorithm

Many real-life signals, such as gravitational wave measurements, biomedical signals, or geophysical data, are strongly non-stationary but can be decomposed into mono-component signals that contain only one active frequency over time. This is made possible thanks to decomposition methods developed in recent years that can handle non-stationary signals. The problem now is how to compute, in an accurate and stable way, the instantaneous frequency, phase, and amplitude of such mono-component signals. Numerous approaches have been developed so far, but they can be unstable in the presence of noise and struggle to capture quick and intrawave changes in frequency. In this work, we present an alternative approach, called the JADE method, which is based on the Dynamic Time Warping algorithm and which we combine with the FIF algorithm to handle and study multicomponent non-stationary signals. We test the robustness of JADE to noise and run comparisons with classical methods used for instantaneous frequency, phase, and amplitude estimation.
Read the full abstract

Statistics Theory

A note on kernel density estimators with optimal bandwidths

We show that the cumulative distribution function corresponding to a kernel density estimator with optimal bandwidth lies outside any confidence interval, around the empirical distribution function, with probability tending to 1 as the sample size increases.

2604.14181 · 31 Mar · 8 pages, 0 figures. Statistical Research Report, Department of Mathematics, University of Oslo, from June 2000, but arXiv'd April 2026. The papers is pubished in essentially this form in Statistics & Probabiity Letters, 2001, vol. 54, pages 153-159, at this url: https://www.sciencedirect.com/science/article/pii/S016771520100027X

General Mathematics

Biharmonic Subdivision on Riemannian Manifolds

This paper introduces a biharmonic interpolatory subdivision framework on Riemannian manifolds. In the Euclidean setting, the six-point Deslauriers-Dubuc stencil is characterised as the unique minimiser of a discrete curvature-variation energy under symmetric six-point support and degree-five polynomial reproduction conditions, linking a classical interpolatory rule to a first-principles fairness criterion. Exact symbol analysis establishes fourth-order smoothness. The construction extends to the two-sphere and the hyperbolic plane via a second-order reduced governing ODE derived from the biharmonic Euler-Lagrange equation on constant-curvature surfaces. This reduced model yields closed-form insertion rules, and proximity analysis confirms that the manifold scheme satisfies the Wallner-Dyn second-order condition, preserving fourth-order smoothness. A hierarchy of biharmonic stencils achieving higher smoothness orders is also described. Numerical experiments demonstrate that the six-point scheme delivers lower fairness energy and smoother curvature profiles than the classical four-point Dyn-Gregory-Levin scheme, while remaining more local and exhibiting less ringing on non-uniform data than the eight-point variant.
Read the full abstract

General Mathematics

A Theory of Scales and Orbit Covers

This paper develops a formal theory of musical scales and their harmonic coverings and introduces orbit covers: coverings obtained by translating a fixed subset across a scale via a group action. Orbit covers generalize familiar constructions, such as the covering of the diatonic scale by tertian triads, and are motivated by the search for a generalized harmonic framework extending common-practice tonality. We model modes as group structures associated with pitch-class sets and scales as torsors, introducing scale covers and, in particular, orbit covers. To each orbit cover we associate a nerve complex encoding its intersection structure and associated topological invariants. We classify triadic orbit covers of heptatonic scales up to affine symmetry and nerve isomorphism. These results support a broader theory of harmonic organization with analytical and compositional applications.
Read the full abstract

Analysis of PDEs · Cross-listed from nlin.SI

Finite-time blow-up of two $(1+1)$D systems rigorously derived from the 3D axisymmetric Euler equations

We study two $(1+1)$-dimensional systems, denoted $(R0)$ and $(Z0)$, that are rigorously derived from the three-dimensional axisymmetric Euler equations in a signed polar formulation on the meridian plane. The main point of view in this revision is that these $(1+1)$D systems are not ad hoc model equations and not merely “symmetry-axis reductions.” Rather, they arise as exact symmetry-axis/apex restrictions of the full $(1+2)$D system~$(E2)$ obtained from 3D axisymmetric Euler, and they already contain the core finite-time singularity mechanism of the full problem. The rev3 geometry is based on the symmetry axes \[ \theta=0,\qquad \theta=\pm \frac{\pi}{2}, \] for which ridge flatness is preserved automatically by the evenness in $(r,z)$. Along these axes, and in particular at the apex $x^2=r^2+z^2=0$, the reduced dynamics closes exactly. This yields two rigorously derived $(1+1)$D systems: the horizontal-axis system $(R0)$ and the vertical-axis system $(Z0)$. The apex trace of these systems reduces further to a closed ODE of Constantin–Lax–Majda type, from which we obtain finite-time blow-up at the coordinate origin. The paper has three main outputs. First, it derives the signed-polar $(1+2)$D subsystem~$(E2)$ from the 3D axisymmetric Euler equations and identifies the exact $(1+1)$D systems $(R0)$ and $(Z0)$ carried by the symmetry axes. Second, it proves finite-time blow-up for the resulting apex dynamics and analyzes the associated convective axis reduction. Third, it derives the exact background–remainder equations and formulates a conditional nonlinear stability mechanism: if a compatible full background exists on $[0,T)$ with the coefficient bounds required by the weighted energy method, then the full solution inherits the same finite-time apex blow-up.
Read the full abstract

2604.01244 · 31 Mar · 32 pages (Compared with revision2, the present version introduces a new rev3 geometry based on symmetry axes $\theta=0,\pm\pi/2$, rigorously derives two exact (1+1)D systems from the 3D axisymmetric Euler equations, proves finite-time apex blow-up through the closed dynamics at x=0). arXiv admin note: substantial text overlap with arXiv:2603.26715

Numerical Analysis

Diffusion models with physics-guided inference for solving partial differential equations

Diffusion models have recently emerged as powerful stochastic frameworks for high-dimensional inference and generation. However, existing applications to partial differential equations (PDEs) predominantly rely on physics-informed training strategies, which tightly couple learning with specific governing equations and limit generalization across problem settings. In this work, we propose a diffusion model with physics-guided inference for solving PDEs, in which the diffusion model is trained using standard data-driven procedures, while physical laws are incorporated exclusively during the reverse inference stage. The reverse diffusion dynamics is guided by a PDE residual energy function, combined with Gaussian smoothing and explicit boundary enforcement, yielding a physically consistent stochastic iteration that is independent of the training process. From a numerical standpoint, the proposed framework can be interpreted as a diffusion-inspired implicit solver that converges to the PDE solution even when initialized from random noise and perturbed by stochastic fluctuations. The method is validated on classical PDE equation such as Poisson, Diffusion, and Burgers equations with varying coefficients. Numerical results demonstrate robust convergence, high accuracy, and strong generalization without retraining, highlighting the proposed framework as a unified alternative to classical numerical solvers and physics-informed neural networks.
Read the full abstract

All authors Yi Bing, Liu Jia, Fu Jinyang, Peng Xiang

Mathematical Physics · Cross-listed from cs.CC

Stable algorithms cannot reliably find isolated perceptron solutions

We study the binary perceptron, a random constraint satisfaction problem that asks to find a Boolean vector in the intersection of independently chosen random halfspaces. A striking feature of this model is that at every positive constraint density, it is expected that a $1-o_N(1)$ fraction of solutions are strongly isolated, i.e. separated from all others by Hamming distance $\Omega(N)$. At the same time, efficient algorithms are known to find solutions at certain positive constraint densities. This raises a natural question: can any isolated solution be algorithmically visible? We answer this in the negative: no algorithm whose output is stable under a tiny Gaussian resampling of the disorder can reliably locate isolated solutions. We show that any stable algorithm has success probability at most $\frac{3\sqrt{17}-9}{4}+o_N(1)\leq 0.84233$. Furthermore, every stable algorithm that finds a solution with probability $1-o_N(1)$ finds an isolated solution with probability $o_N(1)$. The class of stable algorithms we consider includes degree-$D$ polynomials up to $D\leq o(N/\log N)$; under the low-degree heuristic \cite{hopkins2018statistical}, this suggests that locating strongly isolated solutions requires running time $\exp(\widetilde{\Theta}(N))$. Our proof does not use the overlap gap property. Instead, we show via Pitt's correlation inequality that after a random perturbation of the disorder, the number of solutions located close to a pre-existing isolated solution cannot concentrate at $1$.
Read the full abstract

All authors Shuyang Gong, Brice Huang, Shuangping Li, Mark Sellke

Analysis of PDEs

Quantification of ergodicity for Hamilton–Jacobi equations in a dynamic random environment

We study quantitative large-time averages for Hamilton–Jacobi equations in a dynamic random environment that is stationary ergodic and has unit-range dependence in time. Our motivation comes from stochastic growth models related to the tensionless (inviscid) KPZ equation, which can be formulated as Hamilton–Jacobi equations with random forcing. Understanding the large-time behavior of solutions is closely connected to fundamental questions concerning fluctuations and scaling in such growth processes. In this article, we establish, up to slowly varying factors, convergence rates with exponent $1/2$ for the large-time averages of both the solutions and the associated metric problem toward their ergodic limits. Our proof relies crucially on a new almost-Lipschitz regularity theory for the metric problem, which is of independent interest.
Read the full abstract

All authors Xiaoqin Guo, Wenjia Jing, Hung Vinh Tran, Yuming Paul Zhang

Dynamical Systems · Cross-listed from eess.SY

Nonlinear Moving-Horizon Estimation Using State- and Control-Dependent Models

This paper presents a state- and control-dependent moving-horizon estimation (SCD-MHE) algorithm for nonlinear discrete-time systems. The nonlinear dynamics and measurement maps are written exactly in pseudo-linear form using state- and control-dependent coefficient (SCDC) matrices. At each time step, the moving-horizon estimation problem is solved by a sequence of sparse quadratic programs, where the SCDC matrices are refrozen along the trajectory computed by the preceding iteration; the estimator requires no Jacobians and retains the nonlinear model exactly. We show that each quadratic program has a unique solution, characterize the fixed points of the iteration, establish geometric convergence under a contraction condition, and prove that the estimation error of the computed estimate is uniformly bounded under uniform observability, bounded disturbances, a bounded arrival-cost error, and iterates confined to a compact set, for all finite iteration counts. In a quadrotor benchmark with a saturating rangefinder, the altitude RMSE of SCD-MHE is 18 times smaller than that of a nonlinear moving-horizon estimator that solves the nonlinear program and 58 times smaller than that of the extended and unscented Kalman filters, and its per-step computation time is 34 times smaller than that of the nonlinear estimator.
Read the full abstract

Mathematical Physics

Localised Davies generators for (pseudo)differential operators

A classical Davies generator provides a Lindbladian for which the Gibbs state is stationary. Its construction involves precise knowledge of the Bohr spectrum or equivalently state evolution for all times. Recently Chen, Kastoryano and Gilyén proposed a construction involving localisation in time and carried out in the case of finite dimensional Hilbert spaces. The resulting generators are called quantum Gibbs samplers as the corresponding Lindblad evolution is expected to settle to the Gibbs state. In this paper, we show that the localised Davies construction also works for natural classes of unbounded operators, including pseudodifferential operators used in the study of classical/quantum correspondence in Lindblad evolution. Our emphasis is microlocal: we prove that the localised jump operators are themselves pseudodifferential, and hence pseudolocal. The proof involves a novel version of Egorov's theorem valid for all times.
Read the full abstract

Dynamical Systems · Cross-listed from eess.SY

Set-Based Value Function Characterization and Neural Approximation of Stabilization Domains for Input-Constrained Discrete-Time Systems

Analyzing nonlinear systems with stabilizable controlled invariant sets (CISs) requires accurate estimation of their domains of stabilization (DOS) together with associated stabilizing controllers. Despite extensive research, estimating DOSs for general nonlinear systems remains challenging due to fundamental theoretical and computational limitations. In this paper, we propose a novel framework for estimating DOSs for controlled input-constrained discrete-time systems. The DOS is characterized via newly introduced value functions defined on metric spaces of compact sets. We establish the fundamental properties of these value functions and derive the associated Bellman-type (Zubov-type) functional equations. Building on this characterization, we develop a physics-informed neural network (NN) framework that learns the value functions by embedding the derived functional equations directly into the training process. The proposed methodology is demonstrated through two numerical examples, illustrating its ability to accurately estimate DOSs and synthesize stabilizing controllers from the learned value functions.
Read the full abstract

Numerical Analysis

EPO: A Unified Framework for Entropy Stability, Positivity, and Oscillation Suppression

High-order finite volume and discontinuous Galerkin methods are often stabilized by separate nonlinear devices for admissibility, entropy control, and oscillation suppression. This separation hides a simple geometric fact: all three act on the same cellwise candidate state. We propose a general framework (termed EPO) unifying fully discrete entropy stability, positivity/bound preservation, and spurious oscillation elimination. Starting from a candidate update, we scale along the ray anchored at its updated cell average. The admissible-state constraint, the entropy constraint, and the oscillation-suppressing constraint each define an admissibility radius on that ray, and the applied limiter is their minimum. The decisive analytical ingredient is a weak entropy stability at the level of the updated cell average. A two-point Lax–Friedrichs/Riemann-average entropy inequality yields local cell-average entropy budgets, and the same radial scaling mechanism behind Zhang–Shu positivity preservation lifts these weak budgets to strong quadrature-based entropy inequalities. The framework is therefore not a summation-by-parts, split-form, or flux-differencing construction: EPO acts on a candidate finite volume or discontinuous Galerkin update and converts weak average information into fully discrete nodal entropy stability. The construction also works for any prescribed finite family of convex entropy pairs. Each pair yields its own entropy radius, and taking the minimum enforces fully discrete entropy stability for all of them simultaneously. We prove the preservation of cell averages, invariant-set preservation, local and global strong entropy inequalities, stagewise budgets for strong-stability-preserving (SSP) Runge–Kutta methods, an SSP multistep variant that retains the designed high-order temporal accuracy, and extensions on rectangular and unstructured triangular meshes.
Read the full abstract

Analysis of PDEs · Cross-listed from gr-qc

Big bang stability and isotropisation for the Einstein-scalar field equations in the ekpyrotic regime

It has been shown that, in spacetime dimensions $n\geq 3$, that the Kasner-scalar field solutions to the Einstein-scalar fields equations with potential $V_0 e^{-s \phi}$, where $s<s_c=\sqrt{\frac{8(n-1)}{n-2}}$ and $V_0\in \mathbb{R}$, are nonlinearly stable to the past and terminate at a quiescent big bang singularity over the full range of sub-critical Kasner exponents. In particular, the spatially homogeneous and isotropic solutions, the Friedman-Lemaitre-Robertson-Walker (FLRW) spacetimes, to the Einstein-scalar field equations are stable in this sense for $s<s_c$ and $V_0\in\mathbb{R}$. While perturbations of the sub-critical Kasner-scalar field family of solutions, including the FLRW solutions, are asymptotically velocity term dominated (AVTD) near the big bang, they do not in general isotropise near the big bang singularity. Rather, they remain highly anisotropic, even for small perturbations of the isotropic FLRW solutions. In this article, we establish the stability of FLRW solutions to the Einstein-scalar field equations for the potential parameter values $s>s_c$ and $V_0<0$. Such scalar field potentials are known in the literature as ekpyrotic. In particular, we prove that the FLRW solutions to the Einstein-scalar field equations are nonlinearly stable to the past and terminate at a quiescent, crushing AVTD big bang singularity. A distinguishing property of these perturbed spacetimes is that they isotropise towards the big bang.
Read the full abstract

All authors Florian Beyer, David Garfinkle, James Isenberg, Todd A. Oliynyk

Representation Theory

The Drinfeld center of an oligomorphic tensor category

Recently, Harman and the second author introduced a new construction of pre-Tannakian tensor categories based on oligomorphic groups. We develop tools for analyzing the Drinfeld centers of these categories, and compute the center explicitly in a number of cases. In particular, we find several finitely tensor-generated pre-Tannakian categories (including the Delannoy category) that are identified with their own center via the canonical functor; prior to this work, we knew no such examples besides the category of vector spaces.
Read the full abstract

Algebraic Geometry

Local multiplicities for an equivariantly enriched non-transverse Bézout's theorem

We introduce the degree and local degree in equivariant motivic homotopy theory for the purpose of studying equivariant enumerative problems over general fields. Given a finite, tame group scheme $G$ over a field $k$ and an equivariant motivic ring spectrum $E_G$, we define the equivariant motivic degree and a corresponding local degree of a relatively $E_G$-oriented, proper, quasi-smooth morphism of $G$-schemes. We prove a local to global formula expressing the global degree as a sum of local contributions over $G$-orbits. Using these constructions, we define the Euler number of an oriented vector bundle on a quasi-smooth, proper derived stack and show that the Euler number is independent of the choice of section under appropriate hypotheses. In the presence of a finite group action, the equivariant Euler number can be computed as a sum of local equivariant degrees. As an application, we obtain an equivariantly enriched local multiplicity formula for an equivariant non-transverse Bézout theorem, expressing an equivariant intersection number as a sum of local equivariant degrees.
Read the full abstract

Differential Geometry · Cross-listed from physics.comp-ph

Geometry-informed neural atlas for boundary value problems of complex 3D geometries

When three-dimensional bodies contain thin features, non-trivial topology, or scan-derived surfaces, volumetric meshing can become the dominant bottleneck in simulation workflows. We replace this step with a learned geometric representation: overlapping volumetric coordinate charts, each equipped with a neural decoder and Jacobian, trained from point-cloud or level-set data to form a differentiable atlas. Governing equations are pulled back to chart-local reference coordinates via the Piola identity, and local solutions are coupled through multiplicative Schwarz iterations on the overlap graph. Because the atlas is constructed independently of the downstream discretization, one frozen geometric substrate can support fundamentally different solvers (for example, a meshfree physics-informed neural network and a conventional finite-element method) without re-meshing or re-parametrization. Benchmark and verification studies show that the learned atlas preserves expected finite-element convergence behavior and enables both forward and inverse analyses on geometries that would otherwise require solver-specific volumetric meshing.
Read the full abstract

Optimization and Control

A Gradient Sampling Algorithm for Noisy Nonsmooth Nonconvex Optimization

An algorithm is proposed, analyzed, and tested for minimizing locally Lipschitz objective functions that may be nonconvex and/or nonsmooth. The algorithm, which is built upon the gradient-sampling methodology, is designed specifically for cases when objective and generalized gradient values might be subject to bounded, uncontrollable errors. A novel assumption pertaining to such errors in nonsmooth settings is introduced that decouples the noise that may be present due to typical errors in derivative estimates with noise that may be due to misidentification of segments of differentiability of the objective. Similarly to state-of-the-art guarantees for noisy smooth optimization, it is proved for the algorithm that, with probability one, either the sequence of objective values will decrease without bound or the algorithm will generate an iterate at which a measure of stationarity is below a threshold that depends proportionally on the error bounds for the objective function and generalized gradient values. The results of numerical experiments are presented, which show that the algorithm can indeed perform approximate optimization robustly despite errors in objective and generalized gradient values.
Read the full abstract

Dynamical Systems · Cross-listed from eess.SY

Hybrid Energy-Based Models for Physical AI: Provably Stable Identification of Port-Hamiltonian Dynamics

Energy-based models (EBMs) implement inference as gradient descent on a learned Lyapunov function, yielding interpretable, structure-preserving alternatives to black-box neural ODEs and aligning naturally with physical AI. Yet their use in system identification remains limited, and existing architectures lack formal stability guarantees that globally preclude unstable modes. We address this gap by introducing an EBM framework for system identification with stable, dissipative, absorbing invariant dynamics. Unlike classical global Lyapunov stability, absorbing invariance expands the class of stability-preserving architectures, enabling more flexible and expressive EBMs. We extend EBM theory to nonsmooth activations by establishing negative energy dissipation via Clarke derivatives and deriving new conditions for radial unboundedness, exposing a stability-expressivity tradeoff in standard EBMs. To overcome this, we introduce a hybrid architecture with a dynamical visible layer and static hidden layers, prove absorbing invariance under mild assumptions, and show that these guarantees extend to port-Hamiltonian EBMs. Experiments on metric-deformed multi-well and ring systems validate the approach, showcasing how our hybrid EBM architecture combines expressivity with sound and provable safety guarantees by design.
Read the full abstract

Representation Theory

Representation theory of the Gelfand quiver and Harish-Chandra modules for $\mathsf{SL}_2(\mathbb{R})$

In 1970, Gelfand posed the problem of classifying the indecomposable objects in a representation category equivalent to the principal block of Harish-Chandra modules for $\mathsf{SL}_2(\mathbb{R})$; explicit solutions were obtained by Bondarenko, and, independently, Crawley-Boevey. In this article, we give a complete answer to Gelfand's problem from a derived category perspective. We classify indecomposable objects in the bounded derived category of nilpotent representations of the Gelfand quiver in terms of band and string complexes, and determine their images under the derived Auslander-Reiten translation, the sign involution, and the contragredient duality. The four main combinatorial classes are characterized in Lie-theoretic as well as homological terms. For the abelian category of nilpotent representations, we provide projective resolutions, standard homological invariants and explicit representation matrices of all indecomposables. Our approach can be extended to arrow ideal completions of path algebras of skew-gentle quivers.
Read the full abstract

Complex Variables

Harmonic mappings, univalence criteria and a theorem of Lehtinen

The harmonic inner radius $\sigma_H(\Omega)$ of a planar domain $\Omega$ is the largest constant with which a univalence criterion via the Schwarzian derivative holds for harmonic mappings. We show that $\sigma_H(\Omega)\leq\sigma_H(\mathbb{D})\leq 3/2$ for the unit disk $\mathbb{D}$ and for every domain $\Omega$ that omits an open set. This is an analogue of a theorem of Lehtinen in the setting of holomorphic functions. We provide two related univalence criteria for harmonic mappings.
Read the full abstract