arXiv++ Combinatorics

We show that the perfect matching function on $n$-vertex graphs requires monotone circuits of size $\smash{2^{n^{Ω(1)}}}$. This improves on the $n^{Ω(\log n)}$ lower bound of Razborov (1985). Our proof uses the standard approximation method together with a new sunflower lemma for matchings.

We take an order-theoretic approach to circuit (string diagram) syntax, treating a circuit as a partial order with additional input-output structure. We define morphisms between circuits and prove a factorisation theorem showing that these can, in the finite case, be regarded as formalising a notion of syntactical circuit rewrites, with quotient maps in particular corresponding to gate composition. We then consider the connectivity of a circuit, expressed as a binary relation between its inputs and outputs, and characterise the concept lattice from formal concept analysis as the unique smallest circuit that admits morphisms from all other circuits with the same connectivity. This has significance for quantum causality, particularly to the study of causal decompositions of unitary transformations. We close by constructing the circuit characterised by the dual statement.

We study rank-three matroids, known as point-line configurations, and their associated matroid varieties, defined as the Zariski closures of their realization spaces. Our focus is on determining finite generating sets of defining equations for these varieties, up to radical, and describing the irreducible components of the corresponding circuit varieties. We generalize the notion of cactus graphs to matroids, introducing a family of point-line configurations whose underlying graphs are cacti. Our analysis includes several classical matroids, such as the Pascal, Pappus, and cactus matroids, for which we provide explicit finite generating sets for their associated matroid ideals. The matroid ideal is the ideal of the matroid variety, whose construction involves a saturation step with respect to all the independence relations of the matroid. This step is computationally very expensive and has only been carried out for very small matroids. We provide a complete generating set of these ideals for the Pascal, Pappus, and cactus matroids. The proofs rely on classical geometric techniques, including liftability arguments and the Grassmann--Cayley algebra, which we use to construct so-called bracket polynomials in these ideals. In addition, we prove that every cactus matroid is realizable and that its matroid variety is irreducible.

We primarily investigate the properties of characteristic polynomials of semimatroids. In particular, we provide a combinatorial interpretation of their coefficients, generalizing the Whitney's Broken Circuit Theorem. We also prove that the unsigned coefficients of the characteristic polynomial form a unimodal and log-concave sequence, extending the Rota-Heron-Welsh Conjecture to semimatroids. Furthermore, we present convolution identities for the multiplicative characteristic and Tutte polynomials of semimatroids using the Möbius conjugation. Finally, motivated by Kochol's work, we introduce assigning matroids to establish connections among semimatroids, hyperplane arrangements, and graph colorings, with a particular focus on their characteristic polynomials.

The canonical class in the realm of counting complexity is #P. It is well known that the problem of counting the models of a propositional formula in disjunctive normal form (#DNF) is complete for #P under Turing reductions. On the other hand, #DNF $\in$ spanL and spanL $\not\subseteq$ #P unless NL = NP. Hence, the class of functions logspace-reducible to #DNF is a strict subset of #P under plausible complexity-theoretic assumptions. By contrast, we show that two calls to a (restricted) #2DNF oracle suffice to capture gapP, namely, that the logspace many-one closure of the subtraction between the results of two #2DNF calls is gapP. Because #P $\not\subseteq$ gapP, #P is strictly contained between one and two #2DNF oracle calls. Surprisingly, the propositional formulas needed in both calls are linear-time computable, and the reduction preserves interesting structural as well as symmetry properties, leading to algorithmic applications. We show that a single subtraction suffices to compensate for the absence of negation while still capturing gapP, i.e., our results carry over to the monotone fragments of #2SAT and #2DNF. Since our reduction is linear-time, it preserves sparsity and, as a consequence we obtain a sparsification lemma for both #2SAT and #2DNF. This has only been known for kSAT with k $\geq$ 3 and respective counting versions. We further show that both #2DNF calls can be combined into a single call if we allow a little postprocessing (computable by AC0- or TC0-circuits). Consequently, we derive refined versions of Toda's Theorem: PH $\subseteq$ [#MON2SAT]$^{log}_{TC0}$ = [#MON2DNF]$^{log}_{TC0}$ and PH $\subseteq$ [#IMPL2SAT]$^{log}_{AC0}$. Our route to these results is via structure-aware reductions that preserve parameters like treewidth up to an additive overhead. The absence of multiplicative overhead indeed yields parameterized SETH-tight lower bounds.

The configuration of a matroid $M$ is the abstract lattice of cyclic flats (flats that are unions of circuits) where we record the size and rank of each cyclic flat, but not the set. One can compute the Tutte polynomial of $M$, and stronger invariants (notably, the $\mathcal{G}$-invariant), from the configuration. Given a matroid $M$ in which certain pairs of cyclic flats are non-modular, we show how to produce a matroid that is not isomorphic to $M$ but has the same configuration as $M$. We show that this construction applies to a lattice path matroid if and only if it is not a fundamental transversal matroid, and we enumerate the connected lattice path matroids on $[n]$ that are fundamental; these results imply that, asymptotically, almost no lattice path matroids are Tutte unique. We give a sufficient condition for a matroid to be determined, up to isomorphism, by its configuration. We treat constructions that yield matroids with different configurations where each matroid is determined by its configuration and all have the same $\mathcal{G}$-invariant. We also show that for any lattice $L$ other than a chain, there are non-isomorphic transversal matroids that have the same configuration and where the lattices of cyclic flats are isomorphic to $L$.

This paper presents an introduction and expository account of a beautiful, current, and active application of recursions to the computation of resistance distance. Resistance distance, also referred to as effective resistance, is a well-known graph metric that arises naturally by considering a graph as an electrical circuit; heuristically resistance distance measures both the number of paths between two vertices in a graph and the cost of each path. This topic finds applications in a rich array of fields including social, biological, ecological, and transportation networks, chemistry, graph theory, numerical linear algebra, and engineering. A variety of methods are used in the field to determine resistance distance including recursive, mathematical, and graphical techniques. Sequences familiar to the readers of the Fibonacci Quarterly such as the Fibonacci and Lucas sequences appear quite often in results in the literature. Twenty five to forty years ago there were a handful of papers on resistance that appeared in the Fibonacci Quarterly and the Proceedings and recently papers on the subject have appeared again. It is hoped that this introductory expository account will interest readers of the Quarterly to renew interest in this current and active field.

Can algebraic geometry enhance the sharpness, robustness, and interpretability of modern neural reasoning models by equipping them with a mathematically grounded inductive bias? To answer this, we introduce Tropical Attention, an attention mechanism grounded in tropical geometry that lifts the attention kernel into tropical projective space, where reasoning is piecewise-linear and 1-Lipschitz, thus preserving the polyhedral decision structure inherent to combinatorial reasoning. We prove that Multi-Head Tropical Attention (MHTA) stacks universally approximate tropical circuits and realize tropical transitive closure through composition, achieving polynomial resource bounds without invoking recurrent mechanisms. These guarantees explain why the induced polyhedral decision boundaries remain sharp and scale-invariant, rather than smoothed by Softmax. Empirically, we show that Tropical Attention delivers stronger out-of-distribution generalization in both length and value, with high robustness against perturbative noise, and substantially faster inference with fewer parameters compared to Softmax-based and recurrent attention baselines. For the first time, we extend neural algorithmic reasoning beyond PTIME problems to NP-hard and NP-complete problems, paving the way toward sharper and more expressive Large Reasoning Models (LRMs) capable of tackling complex combinatorial challenges in phylogenetics, cryptography, particle physics, and mathematical discovery.

For a graph $G$, we define its great shadow $S(G)$ as a construction that duplicates each vertex $v$ in $G$ and sets this duplicated vertex adjacent to $v$ and all neighbors of $v$. Great graph shadows arise naturally in the routing of diode-and-switch circuits for computer keyboards, and are closely related to the Mycielski operation. These diode-and-switch circuits can be routed on a single-sided printed-circuit board if and only if the corresponding great shadow is planar. In this paper, we characterize all graphs with planar great shadows. Such graphs are always bipartite cactus graphs.

We consider geometric and combinatorial characterizations of equiangular tight frames (ETFs), with the former concerning homogeneity of the vector and line symmetry groups and the latter the matroid structure. We introduce the concept of the bender of a frame, which is the collection of short circuits, which in turn are the dependent subsets of frame vectors of minimum size. We also show that ETFs with $k$-homogeneous line symmetry groups have benders which are $k$-designs. Paley ETFs are a known class of ETFs constructed using number theory. We determine the line and vector symmetry groups of all Paley ETFs and show that they are $2$-homogeneous. We additionally characterize all $k$-homogeneous ETFs for $k\geq 3$. Finally, we revisit David Larson's AMS Memoirs \emph{Frames, Bases, and Group Representations} coauthored with Deguang Han and \emph{Wandering Vectors for Unitary Systems and Orthogonal Wavelets} coauthored with Xingde Dai with a modern eye and focus on finite-dimensional Hilbert spaces.

Matroid theory provides a unifying framework for studying dependence across combinatorics, geometry, and applications ranging from rigidity to statistics. In this work, we study circuit varieties of matroids, defined by their minimal dependencies, which play a central role in modeling determinantal varieties, rigidity problems, and conditional independence relations. We introduce an efficient computational strategy for decomposing the circuit variety of a given matroid $M$, based on an algorithm that identifies its maximal degenerations. These degenerations correspond to the largest matroids lying below $M$ in the weak order. Our framework yields explicit and computable decompositions of circuit varieties that were previously out of reach for symbolic or numerical algebra systems. We apply our strategy to several classical configurations, including the Vámos matroid, the unique Steiner quadruple system $S(3,4,8)$, projective and affine planes, the dual of the Fano matroid, and the dual of the graphic matroid of $K_{3,3}$. In each case, we successfully compute the minimal irreducible decomposition of their circuit varieties.

We study a class of signomials whose positive support is the set of vertices of a simplex and which may have several negative support points in the simplex. Various groups of authors have provided an exact characterization for the global nonnegativity of a signomial in this class in terms of circuit signomials and that characterization provides a tractable nonnegativity test. We generalize this characterization to the constrained nonnegativity over a set $X$ under an additional convexity precondition in the exponential moment space. This provides a tractable nonnegativity test over $X$ for the class in terms of a power cone program. Our proof methods rely on a variant of the convex cone of constrained SAGE signomials (sums of arithmetic-geometric exponentials) and the duality theory.

Due to Veblen's Theorem, if a connected multigraph $X$ has even degrees at each vertex, then it is Eulerian and its edge set has a partition into cycles. In this paper, we show that an Eulerian multigraph has a unique partition into cycles if and only if it belongs to the family $\mathcal{S}$, ``bridgeless cactus multigraphs", elements of which are obtained by replacing every edge of a tree with a cycle of length $\geq 2$. Other characterizing conditions for bridgeless cactus multigraphs and digraphs are provided. Furthermore, for a digraph $D$, we list conditions equivalent to having a unique Eulerian circuit, thereby generalizing a previous result of Arratia-Bollobás-Sorkin. In particular, we show that digraphs with a unique Eulerian circuit constitute a subfamily of $\mathcal{S}$, namely, ``Christmas cactus digraphs".

Almost equitable partitions (AEPs) have been linked to cluster synchronization in oscillatory systems, highlighting the importance of structure in collective network dynamics. We provide a general spectral framework that formalizes this connection, showing how eigenvectors associated with AEPs span a subspace of the Laplacian spectrum that governs partition-induced synchronization behavior. This offers a principled reduction of network dynamics, allowing clustered states to be understood in terms of quotient graph projections. Our approach clarifies the conditions under which transient hierarchical clustering and multi-frequency synchronization emerge, and connects these dynamical phenomena directly to network symmetry and community structure. In doing so, we bridge a critical gap between static topology and dynamic behavior-namely, the lack of a spectral method for analyzing synchronization in networks that exhibit exact or approximate structural regularity. Perfect AEPs are rare in real-world networks since most have some degree of irregularity or noise. We define a relaxation of an AEP we call a quasi-equitable partition at level $δ$ ($δ-$QEP). $δ-$QEPs can preserve many of the clustering-relevant properties of AEPs while tolerating structural imperfections and noise. This extension enables us to describe synchronization behavior in more realistic scenarios, where ideal symmetries are rarely present. Our findings have important implications for understanding synchronization patterns in real-world networks, from neural circuits to power grids.

In a simple graph, a shunt is a symmetry which sends an edge to an incident edge (without fixing their shared vertex). The orbit of this edge under the shunt forms a consistent cycle. The important theorem of Biggs and Conway says that in a dart-transitive graph of valence q, there are exactly q-1 orbits of consistent cycles. These ideas have become a useful tool in the area of graphs symmetries, and generalize easily to consistent walks in graphs which are not simple. These walks are not necessarily cycles, or even circuits. This paper considers these walks and their orbits in the venue of dart-transitive maps and classifies them geometrically.

We study MaxCut on 3-regular graphs of minimum girth $g$ for various $g$'s. We obtain new lower bounds on the maximum cut achievable in such graphs by analyzing the Quantum Approximate Optimization Algorithm (QAOA). For $g \geq 16$, at depth $p \geq 7$, the QAOA improves on previously known lower bounds. Our bounds are established through classical numerical analysis of the QAOA's expected performance. This analysis does not produce the actual cuts but establishes their existence. When implemented on a quantum computer, the QAOA provides an efficient algorithm for finding such cuts, using a constant-depth quantum circuit. To our knowledge, this gives an exponential speedup over the best known classical algorithm guaranteed to achieve cuts of this size on graphs of this girth. We also apply the QAOA to the Maximum Independent Set problem on the same class of graphs.

Recently, the study of circuits and cycles within the homology classes of graphs has attracted considerable research interest. However, the detection and counting of shorter circuits in homology classes, especially the shortest ones, remain underexplored. This paper aims to fill this gap by solving the problem of detecting and counting the shortest cycles in homology classes, leveraging the concept of direction-consistent circuits and extending classical results on Eulerian circuits such as Hierholzer's algorithm and the BEST theorem. As an application, we propose the one-carrier transportation routing problem and relate it to a circuit detection problem in graph homology.

Shi, Walsh and Yu demonstrated that any dense planar graph with certain property (known as circuit graph) contains a large near-triangulation. We extend the result to $2$-connected plane graphs, thereby addressing a question posed by them. Using the result, we prove that the planar Tuán number of $2C_k$ is $\left[3-Θ(k^{\log_23})^{-1}\right]n$ when $k\geq 5$.

The Merino-Welsh conjecture states that for a graph $G$ without loops and bridges we have $$\max(T_G(2,0),T_G(0,2))\geq T_G(1,1).$$ Later Jackson proved that for any matroid $M$ without loop and coloop we have $$T_M(3,0)T_M(0,3)\geq T_M(1,1)^2.$$ The value $3$ in this statement was improved to $2.9242$ by Beke, Csáji, Csikvári and Pituk. In this paper, we further improve on this result by showing that $$T_M(2.355,0)T_M(0,2.355)\geq T_M(1,1)^2.$$ We also prove that the Merino--Welsh conjecture is true for matroids $M$, where all circuits of $M$ and its dual $M^*$ have length between $\ell$ and $(\ell-2)^4$ for some $\ell\geq 6$.

A well-known theorem by François Robert expresses the degenerated character of a synchronous Boolean finite dynamical system, in the case where the associated regulatory graph does not contain any circuit: all states of the system go towards a single fixed point. We present a large family of updating modes of Boolean models with the same particularity.