Search arXiv⌕ Search

arXiv subjects

Mario Kummer

Publications and source records attributed to Mario Kummer.

At least 19 recordsLinked to original sources
arXiv2026

A note on bounded ratios

We prove that the set of bounded ratios $\BR(X)$ on a semialgebraic set $X\subset\R^n_{>0}$ is the convex cone of linear forms that are nonnegative on the tropicalization $\trop(X)$. In particular, it is a rational polyhedral convex cone. For $X$ the set of Lorentzian polynomials with fixed M-convex support, it is the dual to the set of M-convex functions. We record an explicit counterexample to a conjecture of Huang--Huh--Soskin--Wang on the bounded ratios on Lorentzian polynomials. The bounded ratio in the counterexample corresponds to the non-hypermetric clique-web facet $\mathrm{CW}^1_7(1,1,1,1,1,-1,-1)$ of the cut cone on seven vertices.

math.CO↗
arXiv2026

Ulrich sheaves and determinantal representations for higher secant varieties of curves

We show that higher secant varieties of smooth projective curves have symmetric admissible determinantal representations with symmetric Ulrich sheaves of rank one if the embedding is sufficiently ample. For secant varieties of real curves, we give conditions for the representing matrices to be positive definite. This allows us, under mild (conjecturally vacuous) conditions, to represent the convex hull of the curve as a spectrahedron whenever it is a hyperbolicity cone. For secant varieties of rational normal curves, we derive very explicit representations in terms of Littlewood--Richardson coefficients. One key tool that we use are the higher Szeg\H{o} kernels and the higher Scorza correspondences associated to a non-effective theta characteristic on the curve.

math.AG↗
arXiv2026

Lorentzian polynomials and matroids over triangular hyperfields 2: Analytic aspects

Br\"and\'en and Huh showed that Lorentzian polynomials unify Hodge-Riemann relations in combinatorics: their supports are M-convex, and every M-convex set supports a Lorentzian polynomial. Baker, Huh, Kummer, and Lorscheid later proved that, for every $q>0$, the projectivized space $\mathbf{P}\operatorname{L}_J$ of Lorentzian polynomials with support $J$ is homeomorphic to the thin Schubert cell $\operatorname{Gr}^{\mathrm{w}}_J(\mathbb{T}_q)$ of weak representations of $J$ over the generalized triangular hyperfield $\mathbb{T}_q$. We study the quantitative relation between Lorentzian polynomials and representations over triangular hyperfields. For every matroid $M$, we prove that some $q>0$ depending on $M$ satisfies $\operatorname{Gr}^{\mathrm{w}}_M(\mathbb{T}_q)\subseteq\mathbf{P}\operatorname{L}_M\subseteq\operatorname{Gr}^{\mathrm{w}}_M(\mathbb{T}_2)$. Thus $\mathbf{P}\operatorname{L}_M$ lies between two thin Schubert cells, each homeomorphic to it. More generally, for every M-convex set $J$, some $q>0$ depending on $J$ satisfies $\operatorname{N}\operatorname{Gr}^{\mathrm{w}}_J(\mathbb{T}_q)\subseteq\mathbf{P}\operatorname{L}_J\subseteq\operatorname{N}\operatorname{Gr}^{\mathrm{w}}_J(\mathbb{T}_2)$, where $\operatorname{N}$ denotes normalization. We also study $q(M):=\sup\{q>0:\operatorname{Gr}^{\mathrm{w}}_M(\mathbb{T}_q)\subseteq\mathbf{P}\operatorname{L}_M\}$. For $q(n):=q(U_{2,n})$, we prove $q(4)=2$ and $q(5)=\log_2 3$, with matching upper and lower bounds of order $1/n$; hence $q(n)=\Theta(1/n)$, so in particular no universal positive lower bound for $q(n)$ exists.

math.CO↗
arXiv2026

Benchmarks in Leipzig

Between April 1 and May 15, 2026, a group of 49 mathematicians compiled a dataset of research-level mathematics questions with known answers. Most of the work was done during the 3-day workshop *Benchmarks in Leipzig* with 35 participants at the Max Planck Institute for Mathematics in the Sciences in Leipzig, Germany. We present the resulting collection of 100 questions. We evaluated these questions in three stages: a single attempt by five state-of-the-art LLMs, followed by a 20-runs-per-model evaluation with three of these models, and finally a 3-run attempt with two heavy-thinking models. After Stage 1, 41 questions remained completely unsolved; after Stage 2, this count dropped to 16; and we concluded Stage 3 with only 2 unsolved questions. This demonstrates that the mathematical reasoning capabilities of LLMs are becoming impressive.

math.HO↗
arXiv2026

Soohak: A Mathematician-Curated Benchmark for Evaluating Research-level Math Capabilities of LLMs

Following the recent achievement of gold-medal performance on the IMO by frontier LLMs, the community is searching for the next meaningful and challenging target for measuring LLM reasoning. Whereas olympiad-style problems measure step-by-step reasoning alone, research-level problems use such reasoning to advance the frontier of mathematical knowledge itself, emerging as a compelling alternative. Yet research-level math benchmarks remain scarce because such problems are difficult to source (e.g., Riemann Bench and FrontierMath-Tier 4 contain 25 and 50 problems, respectively). To support reliable evaluation of next-generation frontier models, we introduce Soohak, a 439-problem benchmark newly authored from scratch by 64 mathematicians. Soohak comprises two subsets. On the Challenge subset, frontier models including Gemini-3-Pro, GPT-5, and Claude-Opus-4.5 reach 30.4%, 26.4%, and 10.4% respectively, leaving substantial headroom, while leading open-weight models such as Qwen3-235B, GPT-OSS-120B, and Kimi-2.5 remain below 15%. Notably, beyond standard problem solving, Soohak introduces a refusal subset that probes a capability intrinsic to research mathematics: recognizing ill-posed problems and pausing rather than producing confident but unjustified answers. On this subset, no model exceeds 50%, identifying refusal as a new optimization target that current models do not directly address. To prevent contamination, the dataset will be publicly released in late 2026, with model evaluations available upon request in the interim.

cs.CL↗
arXiv2025

Totally real divisors on curves

Since the works of Krasnov and Scheiderer, there has been an interest in studying effective totally real divisors on a curve X defined over a real closed field, i.e., effective divisors supported on the real locus. Scheiderer proved that, for smooth curves over the real numbers with nonempty real locus, each divisor of sufficiently high degree is linearly equivalent to an effective totally real one. The smallest degree N(X) with this property is called the totally real divisor threshold. When the field is non-Archimedean, we obtain a classification of topological types of smooth curves for which N(X) can be infinite. As a consequence, for curves over the real numbers we prove that N(X) cannot be bounded from above only in terms of the topological type, unless the real locus has many connected components. We complement this qualitative result with a quantitative lower bound for N(X), depending on metric properties of the Jacobian and the curve in the Bergman metric. Finally, we relate these metric properties to period matrices of X, expressed in a way compatible with the real structure.

math.AG↗
arXiv2025

Non-negative polynomials on generalized elliptic curves

We study the cone of non-negative polynomials on generalized elliptic curves. We show that the zero set of every extreme ray has dense real points. If a generalized elliptic curve is embedded via a complete linear system, then we show that the convex hull of its real points (taken inside any affine chart containing all real points) is a spectrahedron. On the way, we generalize a result by Geyer--Martens on 2-torsion points in the Picard group of smooth real curves (of arbitrary genus) to possibly singular and reducible ones.

math.FA↗
arXiv2025

Lorentzian polynomials and matroids over triangular hyperfields 1: Topological aspects

Lorentzian polynomials serve as a bridge between continuous and discrete convexity, connecting analysis and combinatorics. In this article, we study the topology of the space $\mathbb{P}\textrm{L}_J$ of Lorentzian polynomials on $J$ modulo $\mathbb{R}_{>0}$, which is nonempty if and only if $J$ is the set of bases of a polymatroid. We prove that $\mathbb{P}\textrm{L}_J$ is a manifold with boundary of dimension equal to the Tutte rank of $J$, and more precisely, that it is homeomorphic to a closed Euclidean ball with the Dressian of $J$ removed from its boundary. Furthermore, we show that $\mathbb{P}\textrm{L}_J$ is homeomorphic to the thin Schubert cell $\textrm{Gr}_J(\mathbb{T}_q)$ of $J$ over the triangular hyperfield $\mathbb{T}_q$, introduced by Viro in the context of tropical geometry and Maslov dequantization, for any $q>0$. This identification enables us to apply the representation theory of polymatroids developed in a companion paper, as well as earlier work by the first and fourth authors on foundations of matroids, to give a simple explicit description of $\mathbb{P}\textrm{L}_J$ up to homeomorphism in several key cases. Our results show that $\mathbb{P}\textrm{L}_J$ always admits a compactification homeomorphic to a closed Euclidean ball. They can also be used to answer a question of Br\"and\'en in the negative by showing that the closure of $\mathbb{P}\textrm{L}_J$ within the space of all polynomials modulo $\mathbb{R}_{>0}$ is not homeomorphic to a closed Euclidean ball in general. In addition, we introduce the Hausdorff compactification of the space of rescaling classes of Lorentzian polynomials and show that the Chow quotient of a complex Grassmannian maps naturally to this compactification. This provides a geometric framework that connects the asymptotic structure of the space of Lorentzian polynomials with classical constructions in algebraic geometry.

math.CO↗
arXiv2025

Periodic Hypersurfaces and Lee-Yang Polynomials

We study periodic measures on $\mathbb{R}^n$ whose Fourier transform is confined to a proper double cone, in the sense of Meyer's notion of lighthouse measures. Lee--Yang polynomials provide a natural family of examples: it follows from the work of Kurasov and Sarnak that the torus zero sets of such polynomials are hypersurfaces supporting directional lighthouse measures. We prove a rigidity theorem showing that, under mild assumptions, this is essentially the only possibility. Any periodic $C^{1+\epsilon}$ hypersurface supporting a directional lighthouse measure must arise as the torus zero set of an essentially Lee--Yang polynomial. The proof is based on the recent classification of one-dimensional Fourier quasicrystals and provides a geometric interpretation of this theory.

math.AG↗
arXiv2025

Representation theory for polymatroids

We develop a theory of representations of (discrete) polymatroids over tracts in terms of Pl\"ucker coordinates and suitable Pl\"ucker relations. As special cases, we recover polymatroids themselves as polymatroid representations over the Krasner hyperfield K and M-convex functions as polymatroid representations over the tropical hyperfield. We introduce and study several useful operations for polymatroid representations, such as translation and refined notions of minors and duality which have better properties than the existing definitions; for example, deletion and contraction become dual operations (up to translation) in our setting. We also prove an idempotency principle which asserts that polymatroids which are not translates of matroids are representable only over tracts that are idempotent in a certain specific sense (in particular -1 = 1). The space of all representations of a polymatroid J, which we call the thin Schubert cell of J, is represented by an algebraic object called {universal tract of J. When we restrict to just the 3-term Pl\"ucker relations, we obtain the weak thin Schubert cell, and passing to torus orbits yields the realization space. These are represented by the universal pasture and the foundation of J, respectively. We exhibit a canonical bijection between the universal tract and the universal pasture, which is new even in the case of matroids, and we show that the foundation of a polymatroid is generated by cross ratios. We also describe a (possibly incomplete) list of multiplicative relations between cross ratios. Thin Schubert cells and realization spaces are canonically embedded in certain tori. Over idempotent tracts, we show that thin Schubert cells contain a canonical torus orbit and split naturally as a product of the realization space with this distinguished torus.

math.CO↗
arXiv2025

Adjoints of Polytopes: Determinantal Representations and Smoothness

In this article we study determinantal representations of adjoint hypersurfaces of polytopes. We prove that adjoint polynomials of all polygons can be represented as determinants of tridiagonal symmetric matrices of linear forms with the matrix size being equal to the degree of the adjoint. We prove a sufficient combinatorial condition for a surface in the projective three-space to have a determinantal representation and use it to show that adjoints of all three-dimensional polytopes with at most eight facets and a simple facet hyperplane arrangement admit a determinantal representation. This includes all such polytopes with a smooth adjoint. We demonstrate that, starting from four dimensions, adjoint hypersurfaces may not admit linear determinantal representations. Along the way we prove that, starting from three dimensions, adjoint hypersurfaces are typically singular, in contrast to the two-dimensional case. We also consider a special case of interest to physics, the ABHY associahedron. We construct a determinantal representation of its universal adjoint in three dimensions and show that in higher dimensions a similarly structured representation does not exist.

math.AG↗
arXiv2024

Two results on the Convex Algebraic Geometry of sets with continuous symmetries

We prove two results on convex subsets of Euclidean spaces invariant under an orthogonal group action. First, we show that invariant spectrahedra admit an equivariant spectrahedral description, i.e., can be described by an equivariant linear matrix inequality. Second, we show that the bijection induced by Kostant's Convexity Theorem between convex subsets invariant under a polar representation and convex subsets of a section invariant under the Weyl group preserves the classes of convex semi-algebraic sets, spectrahedral shadows, and rigidly convex sets.

math.AG↗
arXiv2024

Higher Dimensional Fourier Quasicrystals from Lee-Yang Varieties

In this paper, we construct Fourier quasicrystals with unit masses in arbitrary dimensions. This generalizes a one-dimensional construction of Kurasov and Sarnak. To do this, we employ a class of complex algebraic varieties avoiding certain regions in $\mathbb{C}^n$, which generalize hypersurfaces defined by Lee-Yang polynomials. We show that these are Delone almost periodic sets that have at most finite intersection with every discrete periodic set.

math-ph↗
arXiv2024

Maximal Mumford Curves from Planar Graphs

A curve of genus g is maximal Mumford (MM) if it has g+1 ovals and g tropical cycles. We construct full-dimensional families of MM curves in the Hilbert scheme of canonical curves. This rests on first-order deformations of graph curves whose graph is planar.

math.AG↗
arXiv2024

Three results related to the half-plane property of matroids

We settle three problems from the literature on stable and real zero polynomials and their connection to matroid theory. We disprove the weak real zero amalgamation conjecture by Schweighofer and the second author. We disprove a conjecture by Br\"and\'en and D'Le\'on by finding a relaxation of a matroid with the weak half-plane property that does not have the weak half-plane property itself. Finally, we prove that every quaternionic unimodular matroid has the half-plane property which was conjectured by Pendavingh and van Zwam.

math.CO↗
arXiv2023

Ulrich sheaves, the arithmetic writhe and algebraic isotopies of space curves

We establish a connection between the theory of Ulrich sheaves and $\mathbb{A}^1$-homotopy theory. For instance, we prove that the $\mathbb{A}^1$-degree of a morphism between projective varieties, that is relatively oriented by an Ulrich sheaf, is constant on the target even when it is not $\mathbb{A}^1$-chain connected or $\mathbb{A}^1$-connected. Further if an embedded projective variety is the support of a symmetric Ulrich sheaf of rank one, the $\mathbb{A}^1$-degree of all its linear projections can be read off in an explicit way from the free resolution of the Ulrich sheaf. Finally, we construct an Ulrich sheaf on the secant variety of a curve and use this to define an arithmetic version of Viro's encomplexed writhe for curves in $\mathbb{P}^3$. This can be considered to be an arithmetic analogue of a knot invariant. Namely, we define a notion of algebraic isotopy under which the arithmetic writhe is invariant. For rational curves of degree at most four in $\mathbb{P}^3$ we obtain a complete classification up to algebraic isotopies.

math.AG↗
arXiv2023

Some thoughts and experiments on Bergman's compact amalgamation problem

We study the question whether copies of $S^1$ in $\mathrm{SU}(3)$ can be amalgamated in a compact group. This is the simplest instance of a fundamental open problem in the theory of compact groups raised by George Bergman in 1987. Considerable computational experiments suggest that the answer is positive in this case. We obtain a positive answer for a relaxed problem using theoretical considerations.

math.GR↗