Papers

We propose a machine-learning framework for detecting whether a given matrix is a perturbation of a matrix with a large Jordan block. The proposed model achieves high classification accuracy on synthetically generated, robustly perturbed data and outperforms a classical numerical baseline. Moreover, we demonstrate that the learned model generalizes to several classes of matrices not seen during training. These results suggest that the architecture captures structural properties associated with matrix defectiveness.

Temporal difference (TD) learning with linear function approximation is a core method for policy evaluation. Its classical continuous-time description is an ordinary differential equation (ODE), which captures the asymptotic mean dynamics but neglects stochastic fluctuations determining the error floor. We introduce a stochastic differential equation (SDE) approximation for linear TD(0) under Markovian noise. The resulting model distinguishes the contraction dynamics governed by the projected Bellman operator from the influence of Markovian sampling. As a consequence, the model explains the constant-stepsize error floor through the interaction between Markovian long-run covariance and the contraction geometry of the projected Bellman operator.

We prove a Trudinger-Moser type inequality in fractional Sobolev spaces with singularities on smooth compact sets of codimension $k$, where $1 < k < d$ and $sp = d$. The singular term is given by the inverse $d$-th power of the distance to the submanifold. The proof is based on a fractional Hardy inequality adapted to smooth submanifolds, and we show the sharpness of the constant. We also establish the equivalence of two natural fractional Sobolev spaces vanishing on the singular set.

We develop a minimizing-movement framework for parametric finite element approximations of geometric gradient flows with admissible tangential motion. At each time step, the discrete variational problem combines a metric dissipation term for the normal displacement with a surface Dirichlet energy. The metric determines the normal geometric evolution: the $L^2(\Gamma)$ metric gives mean curvature flow, while the $H^{-1}(\Gamma)$ metric gives surface diffusion flow. Tangential velocity is selected independently through weak constraints on the deformation map. The central structural condition is admissibility, namely, that the identity map satisfies the constraint. This condition keeps the identity map available as a comparison function and yields the natural stability estimate. The framework recovers the classical Barrett–Garcke–Nürnberg (BGN) scheme from the unconstrained formulation and the dual minimal-deformation-rate (MDR) scheme from the MDR constraint. We further introduce two new admissible variants: an admissible BGN scheme and a relaxed MDR scheme. For the resulting fully discrete schemes, we prove existence and uniqueness under natural nondegeneracy assumptions and establish unconditional energy stability. Numerical experiments compare the admissible and classical schemes and illustrate their stability properties and mesh-quality behavior.

We present a numerical method for the forward solution of nonlinear partial differential equations (PDEs) in which Bellman-Kalaba quasilinearization reduces the nonlinear problem to a sequence of linear subproblems, each discretized by collocation onto a trial space that is linear in its parameters and solved by a single direct linear least-squares QR factorization. The trial space, which we term Linear-in-Learnables (LiL), comprises representations whose trainable parameters enter linearly, including random-feature extreme learning machines, spectral polynomial bases, and trigonometric expansions, each implemented as a physics-informed neural network. The method thus replaces the nonconvex gradient-based training that limits standard PINNs with a convex per-step solve. We establish local Newton-Kantorovich convergence of the outer iteration to a residual-limited neighborhood under an explicit smallness condition, with the limiting accuracy governed by the best-approximation residual of the trial space rather than by an optimization tolerance. The method, denoted LiL-Q, is assessed on seven benchmarks spanning scalar nonlinear PDEs (Bratu, viscous Burgers, Buckley-Leverett), coupled systems (plane-strain elasticity and the incompressible Navier-Stokes equations in two and three spatial dimensions), and steady-state Darcy flow with heterogeneous permeability. Across these problems, LiL-Q converges in single-digit outer iterations in most cases, even at the coarsest basis sizes and independent of the parameter count. When the exact solution lies in the span of the trial space, the method recovers it to machine precision in a single solve. On the Navier-Stokes benchmarks, it matches or exceeds published PINN solvers with up to two orders of magnitude fewer trainable parameters, without gradient-based optimization.

This paper studies the probability of substructures occurring in random Latin squares. Our main result states that if $\alpha,\beta>0$ are such that $2\alpha+\beta<1$, then there are positive constants $\delta = \delta(\alpha, \beta)$ and $\Delta = \Delta(\alpha, \beta)$ such that if $P$ is a partial Latin square of order $n$ with $k = k(n)$ non-empty cells occupying at most $\alpha n$ rows and $\beta n$ columns, the probability that a random Latin square of order $n$ contains $P$ lies between $(\delta/n)^k$ and $(\Delta/n)^k$. We apply this result to subsquares in random Latin squares to obtain the first proof of the fact that the expected number of subsquares of order $3$ in a random Latin square of order $n$ is non-vanishing as $n \to \infty$. We are also able to provide the best known asymptotics for the expected number of subsquares of order $a$ in a random Latin square of order $n$ when $2<a=o(n^{1/2})$. Finally, we discuss the implications of our result on other configurations in random Latin squares as well as on completions of partial Latin squares.

The Loewner order on Hermitian matrices is a partial order that compares matrices in terms of positive semidefiniteness. The Loewner order plays a key role in many fields such as optimization, numerical linear algebra, control theory, operator theory, and quantum information. A fundamental difficulty is that two or more Hermitian matrices do not necessarily have a unique minimal upper bound (or maximal lower bound). In this paper, we propose an iterative method to exactly compute a minimal upper bound for any finite collection of $n\times n$ Hermitian matrices. It is shown that the algorithm terminates in at most $n$ iterations. The exactitude of the algorithm is proved using standard results from finite-dimensional linear algebra. A self-contained proof of an algebraic characterization of minimality originally explored by Stott is provided. We illustrate the algorithm in examples and also provide an implementation of the algorithm in Python.

We consider the relationships between the domination number of graph, denoted $\gamma$, and the distance-$2$ domination number, denoted $\gamma_2$, and three parameters that lie between them: the packing number, denoted $\rho$, the lower packing number, denoted $\rho_L$, and the isolation number, denoted $\iota$. There has been recent attention on the question of whether $\gamma/\rho$ is bounded or unbounded for various families of graphs. We consider similar questions for the ratios of the five parameters. In particular we show that, while $\gamma/\rho_L$ is unbounded in trees, it holds that $\iota/\gamma_2$ is less than $2$ for all trees. Further, $\gamma/\rho_L$ is at most $3$ in interval graphs, at most~$4$ in permutation graphs, and at most $5$ in general asteroidal-triple-free graphs. We also show that every tree has a set of vertices that is both isolating and a packing, and characterize trees where $\rho=\rho_L$.

We prove an average-degree lower bound on the independence number of uncrowded uniform hypergraphs. For every fixed $r\geq 2$ and every $\eta>0$, there exists $d_*=d_*(r,\eta)$ such that any uncrowded $(r+1)$-uniform hypergraph $G$ with $n$ vertices and average degree $d\geq d_*$ satisfies \[ \alpha(G)\geq (1-\eta)r^{-1/r}\left(\frac{\log d}{d}\right)^{1/r}n. \] The proof combines a cleaning procedure, which reduces the maximum top-layer degree to the average scale, with a random nibble procedure that repeatedly extracts independent vertices while controlling all lower-order degrees created by the process. After an initial top-layer cleaning, we run a trace nibble. Since the residual hypergraph contains traces of all sizes $2,\ldots,r+1$, we track the maximum degrees in every layer. A binomial-type recurrence for this degree profile yields the stated leading constant.

We show that the von Neumann algebra generated by position operators in a 2-periodic interacting Fock space is a type $II_1$ factor. On the probabilistic side, we prove that the squared position operators have a Marchenko-Pastur distribution with respect to the vacuum state, yielding a natural realization of free Poisson laws within this framework.

We prove a weak version of the classical Hodge theorem on piecewise-algebraic spaces, a class of spaces introduced by Kontsevich and Soibelman in [KS00]. Precisely, we first prove the Poincare lemma that computes singular cohomology as a variant of de Rham cohomology. Then, as a weak Hodge theorem, we naturally embed the singular cohomology into the space of harmonic forms, instead of establishing an isomorphism (which does not hold for those spaces). Our approach in the latter is classical: Sobolev space theory. In addition, we give more detailed proofs for the claims in the appendix to [KS00]. This work is part of a program of extending arithmetic intersection theory to singular spaces. In particular, a type of currents in this singular setup is introduced.

We study the global-in-time continuation, past the singularity, of the smooth, non-isentropic, radially symmetric imploding solutions of the compressible Euler equations recently constructed by Chen, Shkoller, and Vicol. In three space dimensions, for all physically relevant adiabatic exponents $\gamma>1$, we consider the Euler solution that evolves smoothly until an implosion singularity forms at the origin at time $t=0$. We then prove that this solution can be uniquely continued for $t>0$ as a reflected outward-propagating shock, sometimes called a reflected blast wave. For $t>0$, the continuation is a globally forward self-similar weak solution of the Euler equations, selected by the Rankine–Hugoniot conditions and the Lax entropy inequality; it is smooth away from the expanding shock sphere and the spatial origin. The structure at the center of symmetry distinguishes these explosions from the classical Guderley reflected shock. In Guderley's continuation, the reflected blast wave leaves a point vacuum at the origin, where the density vanishes. The solutions constructed here exhibit the opposite behavior: for every fixed $t>0$ the density is unbounded at $r=0$ (though it remains locally integrable), while the pressure stays bounded and the temperature vanishes there.

Channel charting (CC) in real-world coordinates is a recently proposed self-supervised machine learning method that maps high-dimensional channel state information (CSI) to user equipment (UE) position. In this paper, we extend CC to also estimate UE orientation, which can further assist tasks such as beamfinding, precoding, and beam- and cell-assignment. To this end, we propose a novel orientation triplet loss that accounts for angle periodicity and an alignment loss that embeds estimated orientations in real-world coordinates in a self-supervised fashion. Using real-world CSI measurements from a standard-compliant 5G NR system, we demonstrate that the proposed method achieves position and orientation estimation accuracy close to that of supervised approaches trained with ground-truth labels.

Channel state information (CSI)-based neural positioning learns a mapping from CSI measurements to user equipment (UE) positions using neural networks. However, most existing performance evaluations utilize randomly partitioned train/test CSI-dataset splits, which fail to reflect the generalization requirements of practical deployments and present optimistic results. In this paper, we study the spatial and temporal generalization of neural positioning with standard-compliant Wi-Fi and 5G NR systems for three real-world CSI datasets acquired in indoor and outdoor environments. We assess generalization with two different architectures, a conventional multilayer perceptron (MLP) and a novel transformer architecture, to unseen spatial regions, unseen UE trajectories, and CSI measurement campaigns separated by one week. Our experiments show that both architectures generalize well in space and time, and the proposed transformer consistently outperforms the MLP in positioning accuracy while requiring fewer model parameters.

A maximal eventually different family is called $n$-productive if the product family $\mathcal{F}^n$ is still maximal. We construct closed $n$-productive families separating these strengthenings of maximality at every $n \geq 1$. Furthermore, we show how to force and construct an even stronger type of $\mathcal{I}_0$-productive family and discuss the relation of productivity to Van Douwen families.

We compute explicitly the infimum of the norms of $W^{1,2}$-extension operators for planar sectors and exhibit an extension operator attaining this infimum. This solves an open problem posed by Maz'ya.

We study the action principle for space-times whose boundary is singular. We suggest that it is natural to treat the singularity as a free boundary, where the variation is unconstrained. Demanding that the action is stationary under such free variations then implies certain (on-shell) boundary conditions at the singularity. We derive these boundary conditions for the case of Einstein gravity coupled to matter and show that, when applied to an initial spacelike singularity, they exclude Kasner-like or BKL space-times, but admit conformally regular space-times (including FLRW models) sourced by fluids satisfying $0 \leq P < \rho$. For standard hot big bang FLRW cosmologies, the admissible linear (scalar, vector, tensor) perturbations satisfy reflecting boundary conditions at the bang, in agreement with large-scale cosmological observations.

Modern AI systems increasingly solve a task not with a single model call but with several imperfect agents working together: some propose pieces of a solution, others verify them, and the results are combined. These systems often outperform any single model, yet it is rarely clear why they succeed or when they will fail. We model such a system as message passing on a sparse graph, the structure that underlies low-density parity-check (LDPC) codes, and extend the density-evolution machinery of coding theory to this richer setting. In our model a task is a set of coupled binary subclaims, and an agent architecture is a sparse, role-typed factor graph whose check nodes are noisy Boolean verifier nodes, each computing a local Boolean function of the subclaims it touches. Three distinct failure modes, all modeled as erasures (an agent abstaining, a verifier returning no usable output, and a message lost between two agents), propagate as the agents exchange set-valued messages. The check agents combine these messages by a single logical-forcing rule that specializes to XOR, AND, OR, implication, and Horn constraints. This is more than a relabeling of LDPC theory: the verifier functions are nonlinear and value-asymmetric, and the three failure modes do not reduce to a single effective channel, so they require new threshold, finite-length, and converse results rather than a direct reuse of parity-check density evolution. We prove a density-evolution theorem that predicts the asymptotic fraction of unresolved subclaims on random role-typed architectures, with an extension to deterministic, locally tree-like graph sequences. The XOR case recovers the classical LDPC recursion on the binary erasure channel (BEC); the AND case exposes an asymmetry between positive and negative verifier certificates.

The Chebfun software system is a Matlab extension based on representations of (piece-wise) smooth one-variable functions by expansions in Chebyshev polynomials. One of Chebfun's attractive features is the ability to provide solutions to nonlinear boundary value problems (BVP) with accuracy close to the machine precision. This is done by the chebop class which provides automatic solutions by performing linearizations of the nonlinear BVP with a Newton method in function spaces and automatic differentiation. A drawback of chebop automatic approach is the possible lack of convergence of the Newton method if the initial guess is not close enough to the exact solution. An explicit functional Newton-Frechet linearization done for each particular shape of the differential operator (i.e. without automatic differentiation) proves to be more robust than the chebop class and allows an enlargement of the range of convergence. Another alternative is the implicit L-scheme (quasi-Newton approach with derivatives replaced by suitable positive constants L), with a much simpler implementation and globally convergent. Mixed methods L-scheme/chebop and L-scheme/Frechet may be used to speed up the computation while preserving the robustness and global convergence of the L-scheme. While chebop is the easiest way to solve BVPs, provided that it converges, the alternative approaches overcome the convergence issue, yielding accurate solutions to a wide class of steady-state one-dimensional problems governed by Richards equation. Chebfun2 and Chebfun3, which currently cannot solve BVPs, provide efficient tools for accuracy assessments of the unsteady solutions in one or two spatial dimensions obtained by classical discretization schemes.

Inspired by \cite{CCP}, we investigate a delayed leader-follower Cucker-Smale model describing the collective dynamics of interacting agents subject to communication lags. We first study the particle dynamics and establish sufficient conditions ensuring the emergence of asymptotic flocking. Our analysis shows that velocity alignment and bounded spatial dispersion persist despite the presence of delays and heterogeneous interactions between leaders and followers. We then derive and analyze two continuum descriptions of the system. In the first regime, the number of leaders is kept fixed while the number of followers tends to infinity, leading to a hybrid particle-kinetic model. In the second regime, both populations become infinitely large, yielding a fully kinetic delayed leader-follower model. For both mean-field formulations, we prove global existence, uniqueness, and Wasserstein stability of measure-valued solutions. These results provide a rigorous mathematical framework for the study of collective dynamics with leadership and memory effects and establish a bridge between delayed flocking models and their continuum counterparts.

We establish weighted recursions for the coefficients of Ramanujan's third order mock theta functions $f$ and $\omega$. Specifically, we apply a holomorphic projection operator to vector-valued Rankin-Cohen brackets of completed mock theta series and their shadows. By employing a vector-valued framework, we exploit the vanishing of certain spaces of vector-valued cusp forms. Our proof is AI-assisted and prioritizes accessibility, allowing for straightforward customization and replication within the broader research community.

In this short note, using the Sarkisov link between a prime Fano threefold $V_{22}$ of degree $22$ and the quintic del Pezzo threefold $V_5$, we prove that the Hilbert scheme of rational quartic curves in $V_{22}$ admits a generically $2$-to-$1$ rational map onto the projective space $\mathbb{P}^4$.

We characterise the density of the positive rank one subspace of a masa-bimodule in terms of its support. We prove that strongly reflexive masa-bimodules form a Boolean lattice under naturally defined operations. We examine the lattice-theoretic properties of the class of strongly reflexive masa-bimodules that are also operator systems and some natural subclasses thereof, and provide a topological description of the lattice operations in the case the masa-bimodules arise from closed subsets of a locally compact group.

Given a hyperbolic group $G$ and a quasiconvex subgroup $Q$, we define a \emph{branched cover of $G$ along $g$}, which is a hyperbolic group $H$ with a certain map into $G$. This builds on recent work on drilling hyperbolic groups and generalizes the case where $G$ is the fundamental group of a closed hyperbolic $3$-manifold $M$, $Q \cong \mathbb{Z}$ is represented by an embedded geodesic loop $\gamma$, and $H$ is the fundamental group of a branched cover of $M$ with branching locus $\gamma$. We show that certain deepness assumptions on Dehn fillings induce branched covers, providing many examples of such branched covers. Some additional assumptions imply these branched covers have boundary $S^2$, which may hold interest for the Cannon Conjecture.

In this paper, we propose a general-purpose multi-dimensional symbol construction for computing an arbitrary symmetric function with digital over-the-air computation (OAC) and discuss the practical aspects of coherent aggregation. For our first contribution, we discuss the categorical representation of a symmetric function. By using this representation and leveraging the sufficiency of the histogram to evaluate a symmetric function, i.e., inspired by type-based multiple access (TBMA), we introduce a general approach to design a single set of OAC symbols to compute any digital function. For our second contribution, we use a comprehensive platform based on low-cost nodes that maintain synchronization in time, frequency, phase, and amplitude via a trigger mechanism, enabling coherent OAC experiments without Global Positioning System (GPS) or cable-based synchronization. Using measurements from the platform, we characterize the phase and amplitude statistics of the composite channel to derive a realistic impairment model for coherent OAC. Through a comprehensive analysis, we demonstrate the effectiveness of the proposed scheme under impairments captured by the proposed model

The generalized Krylov subspace (GKS) method is an effective projection technique for large-scale Tikhonov regularization with a general regularization matrix. As the subspace expands, however, two computational bottlenecks limit scalability: the thin QR factorizations of the tall projected matrices formed by the forward operator and the regularization matrix applied to the basis, and the full reorthogonalization of each new basis vector against all previous columns. We propose a sketched variant, named sGKS, that addresses both bottlenecks. The QR factorizations are performed on compressed matrices of much smaller row dimension, maintained incrementally via rank-one updates. Moreover, we observe that explicit reorthogonalization can be skipped entirely without compromising the quality of the approximation subspace, since no step of GKS relies intrinsically on the orthogonality of the basis. The resulting algorithm is independent of the choice of sketching operator and preserves the approximation quality of the original method: we show that, in the absence of sketching in the projected solve, sGKS produces iterates identical to those of standard GKS, and that the sketched projected solve delivers quasi-optimal residual norms controlled by the embedding quality. For more challenging problems where the loss of basis orthogonality becomes significant, we show that incorporating a small number of iterative refinement steps in the projected solve restores the spectral properties of the basis and recovers the full accuracy of the unsketched method. Numerical experiments on image deblurring, X-ray computerized tomography, seismic travel-time tomography, and dynamic computerized tomography demonstrate that sGKS matches the reconstruction quality of standard GKS while significantly reducing per-iteration costs and overall wall-clock time.

Let $G$ be a finite connected weighted graph and let $R$ be its effective-resistance matrix. For every nonempty vertex set $S$, we factor the cofactor sum and determinant of the principal resistance submatrix $R[S]$ into an enumerative term and a boundary potential-theoretic term. If $\tau(G)$ is the weighted spanning tree enumerator and $\kappa_G(S)$ is the weighted enumerator of $S$-rooted spanning forests, then \[ \cof R[S]=(-2)^{|S|-1}\kappa_G(S)/\tau(G). \] After Kron reduction to $S$, with reduced Laplacian $K=L^S$, $Q=K^+$, and $q=\diag(Q)$, the remaining normalized factor is \[ \det R[S]/\cof R[S] =\frac{2}{|S|}\tr Q+\frac12 q^{\mathsf T}Kq. \] The cofactor factor is a principal specialization of known resistance-minor identities; the contribution here is the boundary/Kron-reduction factorization and local radius calculus. Equivalently, the normalized factor is the maximum of $u^{\mathsf T}R[S]u$ over all $u\in\R^S$ satisfying $\one^{\mathsf T}u=1$. This optimization viewpoint yields monotonicity under enlargement of $S$, an exact one-point update formula, and a support criterion for equality. Small star examples show that the resulting set function is neither submodular nor supermodular in general.

A foundational theorem of Ajtai, Komlós, Pintz, Spencer, and Szemerédi asserts that every $n$-vertex $k$-uniform uncrowded hypergraph (i.e., with girth at least $5$) of maximum degree at most $\Delta$ contains an independent set of size $cn(\log\Delta/\Delta)^{\frac{1}{k-1}}$ for some absolute constant $c>0$. Improving an earlier bound of Frieze and Mubayi, Iliopoulos showed that every such hypergraph has list chromatic number at most $(1+o(1))(k-1)(\Delta/\log\Delta)^{\frac{1}{k-1}}$. Determining the optimal leading constant in this bound remains a major open problem. A natural target, suggested by the coupon-collector heuristic, is $(k-1)^{\frac{1}{k-1}}$, which also corresponds to the shattering threshold for coloring sparse random hypergraphs. Our main result shows that this target is attained: for every integer $k\ge2$, every $k$-uniform uncrowded hypergraph of maximum degree at most $\Delta$ satisfies: $$\chi_\ell(\mathcal{H})\le(1+o(1))\Big((k-1)\frac{\Delta}{\log\Delta}\Big)^{\frac{1}{k-1}}.$$ In particular, this confirms a 2019 conjecture of Molloy for uncrowded hypergraphs. Building on this result, we prove that every $k$-uniform linear hypergraph of maximum degree at most $\Delta$ has chromatic number at most $(c_k+o(1))(\Delta/\log\Delta)^{\frac{1}{k-1}}$, where $c_k\to1$ as $k\to\infty$, thereby proving a conjecture of Verstraëte and Wilson on the independence number of linear hypergraphs in the stronger form of colorings. Our techniques also yield a palette sparsification theorem for uncrowded hypergraphs, partially answering a question of Casselgren and Eriksson.

In an asymmetric shelf shuffle, a deck of $n$ cards is dealt sequentially from the bottom and assigned one of the $m$ shelves uniformly at random. The card is placed at the top of the assigned shelf with probability $p$, and at the bottom of the assigned shelf with probability $(1-p)$. Analysis of the shelf shuffle has gained much attention recently, and the case $p=1/2$ was first treated by Diaconis–Fulman–Holmes [Ann. Appl. Prob. 23 (2013), no. 4, 1692–1720]. In this paper, we extend the analysis of the shelf shuffle to general $p\in (0, 1)$. In particular, we study the distribution of cycles, cycle lengths, number of descents, number of valleys, number of inversions, and the RSK shape of a permutation obtained from an asymmetric shelf shuffle. Our results extend the analysis of Diaconis–Fulman–Holmes to arbitrary $p$. Furthermore, our analysis of the distribution of descents and inversions is new even for $p=1/2$.

Let $a\ge2$, let $p\in\mathbb{Z}[n]$ be eventually increasing and eventually non-negative, and let $\lambda_n=a^{p(n)}+f(n)>0$, where $f(n)\in\mathbb{Z}$. We prove, using Bourgain's bounded entropy criterion, that if $\log(1+f(n)a^{-p(n)})$ is eventually non-zero and decays geometrically in absolute value, then $(\lambda_n)$ is neither $L^\infty$-Khintchin nor $L^1$-Khintchin. In particular, for every $c\in\mathbb{Z}\setminus\{0\}$, every positive tail of $(a^{p(n)}+c)_{n\ge1}$ is non-Khintchin. The same conclusion applies to the standard examples $a^n+c$, $a^n+b^n$, and, whenever eventually positive, $a^n-b^n$, with $a\neq b$. Thus these perturbations of geometric powers lie on the unstable side of the Khintchin problem. This gives a negative answer, in the translated-power case, to the question of Fan–Fan–Queffélec–Queffélec on the stability of translated powers.