arXiv++ Combinatorics

Dunshee and Ellingham recently showed that seven natural properties of a cellularly embedded graph form a Fano-plane framework. We establish an analogous framework for binary delta-matroids. For a binary delta-matroid $D$ and $τ$, let $Z_3(D,τ)$ denote its associated binary tight $3$-matroid. The six outer points are represented by evenness or bipartiteness of $D$ and its global vertex-flip transforms. For the seventh point, we call $D$ $Z_3$-bipartite when every circuit of $Z_3(D,τ)$ has even cardinality. We show that the satisfied properties are precisely the nonzero vectors of a subspace of $\Ftwo^3$. For ribbon-graphic delta-matroids, $Z_3$-bipartiteness is equivalent to bipartiteness of the medial graph, so the construction recovers the Fano-plane framework for embedded graphs.

An Euler circuit of a graph is a closed path that visits every edge of the graph exactly once. Euler circuit and circuit decomposition problems can also be formulated for higher dimensional simplicial complexes. An Euler k-circuit in K is a cyclic sequence of vertices v_1...v_n such that every k+1 adjacent terms { v_i,v_{i+1},...,v_{i+k} } (indexed modulo n) form a k-simplex, and every k-simplex of K appears exactly once in the sequence v_1v_2...v_n(v_1v_2...v_k). We investigate the 2-circuit decomposition problem for triangulated closed compact surfaces. For an orientable triangulated surface we use interior angles of paths to define an obstruction that lives in the first cohomology of the surface. It vanishes if and only if the surface has a 2-circuit decomposition. We also show that a non-orientable triangulated surface has a 2-circuit decomposition if and only if its orientable 2-fold cover does.

We characterize four groups of Clifford circuits for Calderbank--Shor--Steane (CSS) codes that are relevant to fault-tolerant logical operations. First, we show that every code-preserving Clifford circuit is a product of Z-diagonal circuits, composed of S and CZ gates, and their X-basis analogues. Second, we define the two-fold transversal group, generated by depth-one two-local code-preserving circuits, and show that each of its elements can be expressed as a product of layers consisting of either Z-diagonal, X-diagonal, or CNOT gates. As a corollary, every transversal gate is a product of three transversal diagonal circuits; for connected non-self-dual codes, two such circuits suffice. We further show that every code-preserving automorphism circuit, consisting of single-qubit Clifford gates and permutations, has a normal form comprising a Hadamard layer, a permutation, and two diagonal circuits. We also define a two-fold automorphism group, in which a depth-one two-local circuit may be code-preserving up to a permutation, and show that its logical image can be larger than that of the two-fold transversal group. For 136 CSS codes, we provide explicit generators and determine the logical image of the two-fold transversal group. We find 78 codes whose full logical Clifford group is generated by two-fold-transversal circuits, including codes of distances 3, 4, 5, 6, 8, and 12, with respective rates $2/5$, $3/4$, $1/9$, $1/5$, $2/5$, and $3/56$. We construct three families of CSS codes from bipartite grids, cut-complements, and quadrics, many of which realize the full logical Clifford group in this way. More generally, the induced logical group can be large even when it is not full logical Clifford group: it has order at least $460\,800$ for the gross code and roughly $10^{26}$ for a clustered-cyclic code.

For a matroid $M$, a join is a set $J\subseteq E(M)$ that meets every circuit $C$ in at most $|C|/2$ elements. Let $μ(M)$ denote the maximum size of a join. Motivated by Frank's min--max theorem for graphic matroids, we compare $μ(M)$ with an ear-decomposition parameter $η(M)=(r(M)+\varphi(M))/2$, where $\varphi(M)$ is the minimum number of even lobes in an ear decomposition of $M$. Frank's theorem implies $μ(M)=η(M)$ for connected graphic matroids. Here we study how far this equality extends beyond graphic matroids. We show that the exact equality does not hold in general: it already fails for cographic matroids, hence within the binary class. Furthermore, the class of matroids satisfying $μ(M)=η(M)$ is not minor-closed, thus there is little hope for a forbidden minor characterization. We also prove that computing a maximum join is NP-hard for cographic matroids, hard to approximate within a factor of $519/520$, and NP-hard for sparse paving matroids given by their list of bases. Despite these negative results, we show that the two parameters remain quantitatively comparable in several natural classes. We prove comparison bounds for binary, paving, cographic, and arbitrary connected matroids. In particular, using Seymour's decomposition theorem, we combine the equality for graphic matroids, the bound for cographic matroids, and a direct analysis of $R_{10}$ to obtain $η(M)\leq 6μ(M)-2$ for every regular matroid $M$.

We prove a conjecture of Regts and Sevenster: a complex-valued graph parameter $f$ with $f(\varnothing)=1$ has exponentially bounded edge-connection rank if and only if it is a mixed partition function; moreover, the model may be chosen with its numbers of even and odd colours explicitly bounded in terms of the rank bound. From $f$ we construct a connection category, a rigid symmetric $\mathbb{C}$-linear monoidal category whose morphism spaces have the connection ranks as dimensions and whose trace pairings are nondegenerate. The rank hypothesis forces moderate tensor growth, and a recent theorem of Etingof and Penneys then shows that every nilpotent endomorphism has trace zero; together with the nondegeneracy of the trace pairing, this makes the category semisimple, and a theorem of Deligne provides a faithful symmetric tensor functor to finite-dimensional super vector spaces. We then identify the resulting super tensor network with the Regts-Sevenster model exactly, viz. with its Eulerian-subgraph expansion and its sign of $-1$ for every fermionic circuit. An appendix gives an independent and direct proof of the nilpotent-trace step, showing that in a rigid symmetric $\mathbb{C}$-linear category with $\mathrm{End}(\mathbf{1})=\mathbb{C}$, exponentially bounded endomorphism growth makes the trace zeta function of every endomorphism rational, with explicit degree bounds.

The cycle double cover conjecture of Szekeres and Seymour, the proof of which was recently announced by OpenAI, states that every bridgeless graph has a collection of cycles covering every edge exactly twice. We study the counting version of this statement for cubic graphs, where we count circuit double covers --- collections of circuits (connected 2-regular subgraphs) covering every edge twice. We show that every 2-edge-connected 3-edge-colorable cubic graph on $n$ vertices has at least $2^{n/2-1}$ circuit double covers, matching our previously conjectured general lower bound. For every 3-edge-connected cubic graph with girth at least 16 we show a weaker exponential lower bound on circuit double covers. For both of these results we use the same system of linear equations used by OpenAI in their proof, however, we provide additional combinatorial interpretation. We characterize planarity of a cubic graph by solvability of this system of equations for arbitrary nowhere-zero $\mathbb Z_2^k$-flow. We give a condition on the flow that is equivalent to existence of a 5-cycle double cover.

We prove Sabidussi's compatibility conjecture. Let $G$ be a finite connected multigraph in which every vertex has even degree and the minimum degree is at least four, and let $T$ be a closed trail that traverses every edge exactly once. The edges of $G$ can be partitioned into circuits (connected 2-regular subgraphs) so that no circuit contains the two edges used consecutively anywhere in $T$. In fact, the edges can be four-coloured so that every such pair receives two different colours and the subgraph formed by the edges of each colour has even degree at every vertex. Formalization in Lean 4 is also available in the author's github.

We study tropical circuits with scalar multiplication gates, that is, algebraic circuits whose gates implement $\max$, $+$, or multiplication with a positive constant. For such circuits, we prove exponential size lower bounds for computing maximum weight directed spanning trees and maximum weight bipartite perfect matchings. As a corollary, we obtain an exponential size separation between monotone and non-monotone maxout neural networks, which generalize the popularly used ReLU neural networks. One conclusion from this is that neural network models with enforced convexity constraints, such as input-convex neural networks (ICNNs), sometimes need to be exponentially larger than their unrestricted counterparts in order to express the same functions.

This paper introduces, analyzes, and applies variants of the enumeration framework symmetric lexicographic symmetric-subset reverse search for the enumeration of symmetric feasible subsets of a finite set up to symmetry. The framework is implemented in detail for three applications: cocircuits, circuits, and triangulations of int configurations. There are two new methods presented and analyzed to check the lexicographic minimality of a subset in its orbit: the critical-element method and the modified switch-table method. Moreover, new application-dependent methods to reduce the number of necessary enumeration nodes are introduced: rank-pruning for cocircuits and lex-pruning for triangulations. With a C++-implementation of the ideas in the software package TOPCOM, in all three applications known benchmarks can be computed faster by a large margin. The following new numbers could be computed for the first time (among others): the number of cocircuits of the 9-cube, the number of circuits of the 8-cube, and the number of all triangulations of the product of a 5- and a 3-simplex, as well as the number of all triangulations of a point configuration in dimension six with 17~points with disconnected flip-graph (constructed by Santos). Moreover, for Santos's triangulation it has computationally been checked that its flip-graph component is indeed purely non-regular. Furthermore, in another instance in dimension five with 26 points (also constructed by Santos), a flaw has been detected: Santos's triangulation can be heuristically flipped to a regular triangulation in the original point configuration. In a mildly modified version of the point configuration, the heuristics cannot flip Santos's triangulation to a regular triangulation anymore.

Interval graphs are a special class of chordal graphs, and hence have connections to commutative algebra via Fröberg's theorem that characterizes linear resolutions of squarefree quadratic ideals. In recent years, several hypergraphic analogues of interval and chordal graphs have been proposed, in part as an effort to extend Fröberg's theorem to ideals generated in higher degree. In this paper, we study two such classes from the literature, cointerval hypergraphs and underclosed complexes, and show that they are in fact equivalent up to complementation. We then consider their place in the broader theory of higher-dimensional chordality, proving that an underclosed clutter is chordal in the sense of Woodroofe. As a consequence, we answer a question of Dochtermann and Engström by showing that the associated Alexander dual complexes are vertex decomposable, implying that the corresponding circuit ideals have linear quotients. We furthermore show that these dual complexes have shellings induced by their underclosed vertex orders.

We introduce an approach to distinguishing isomorphism types of graphs based on vector spaces of polynomials that are set-wise invariant under permutations ("separating modules," which are representations of the symmetric group), inspired by the Geometric Complexity Theory approach to separating complexity classes (Mulmuley & Sohoni, SIAM J. Comput., 2001). We characterize the power of this method for distinguishing non-isomorphic graphs under several different complexity measures: - We show that separating modules of "support-degree" $k$ (each monomial touches at most $k$ vertices) are equivalent to the counts of $O(k)$-vertex subgraphs. This is strictly weaker than $O(k)$-dimensional Weisfeiler--Leman (Fürer, ICALP '01). - We show that separating modules of symmetric circuit size $n^{Θ(k)}$ are equivalent to $Θ(k)$-WL. This generalizes and strengthens a result of Dawar & Wilsenach (CSL '18; ICALP '20; ACM Trans. Comput. Log., 2022; Theory Comput., 2025): they proved one direction of this equivalence for invariant polynomials; we generalize to separating modules and prove both directions. - When considering only the multiplicities of separating modules (as was proposed in GCT by Mulmuley & Sohoni, ibid., rather than the polynomials themselves), we show that two graphs are separated by multiplicities if and only if their automorphism groups have different cycle indices. The latter result is notable in the analogy with GCT, as it is the only result we are aware of in which the multiplicity approach to separating isomorphism types of objects has been given an "intrinsic" characterization in terms of the objects themselves. We use this to show that for graphs, multiplicity obstructions are stronger than occurrence obstructions. We also connect invariant polynomials to the Graph Reconstruction Conjectures and Forman's "invariants of finite type" (Adv. Math., 2004).

Motivated by the dimension-bound part of a problem of Bukh, we study Veronese circuits: how large can the span of $t$ linear forms be if their $m$-th powers are minimally linearly dependent? We prove the sharp finite dimension bound \[ \dim L\leq \frac{t+m-2}{m}. \] Here $\ell_1,\ldots,\ell_t$ are nonzero homogeneous linear forms over a field of characteristic zero, the powers $\ell_1^m,\ldots,\ell_t^m$ form a circuit, and $L=\Span\{\ell_1,\ldots,\ell_t\}$. Rational-normal-curve configurations attain equality for infinitely many pairs $(t,\dim L)$; in particular, the affine bound itself is sharp and the optimal leading constant in Bukh's question is $1/m$. The proof uses a coding-theoretic translation: the coefficient row space of the powers is the $m$-th Schur power of the coefficient code, and the minimality hypothesis makes this Schur power a full-support hyperplane to which the Schur-product Kneser theorem of Mirandola and Zémor applies. The same method yields flat-concentration and interpolation criteria, a Cayley--Bacharach lower bound, Segre--Veronese and positive-characteristic variants, and Hilbert-function constraints for equality and near-equality in Veronese circuits.

A biclique decomposition of a graph is a partition of its edges into complete bipartite subgraphs. We consider graphs whose vertices can be ordered such that the neighborhood of every vertex is the union of a sublinear number of intervals. We observe that these graphs admit compact representations in the form of biclique decompositions of small size. Here, the size of a decomposition is measured as the sum of the number of vertices of its bicliques. Combining this result with the existence of suitable vertex orderings for graphs of low neighborhood complexity, as proven by Welzl in 1988, we recover and extend several known results up to logarithmic factors. These results include upper bounds on the Zarankiewicz problem, matrix multiplication, quantum circuit complexity, and shortest path algorithms in ``well-structured'' instances.

Let $M$ be a simple connected binary matroid, and let $e$ and $f$ be distinct elements of $M$. It is well known that, when the only circuits containing $e$ are spanning, $M$ is a circuit with at least three elements. This paper proves that if every circuit containing $\{e,f\}$ is spanning, then the canonical tree decomposition of $M$ is a path in which each vertex is labeled by a circuit, a copy of $U_{1,3}$, or a binary spike having one non-tip element deleted.

The distance-reduction property for a generating set, i.e., a Markov basis, of a toric ideal is a condition that ensures tight connectivity of its fibres. In this paper, we study the distance-reduction property for toric ideals of graphs and move on to explore the relationship between the distance-reduction property and the bouquet structure of homogeneous toric ideals, which includes the class of toric ideals of graphs. For toric ideals of graphs which are complete intersection, we show that the minimal Markov bases are distance-reducing if and only if they distance-reduce the circuits of the ideal. We then consider how the distance-reduction properties interact with the bouquet structure of the toric ideal. Bouquets are a combinatorial structure that capture the essential combinatorial information of the toric ideal. Under the condition of homogeneity, we show that, for toric ideals with the same bouquet structure and signature, the distance-reduction properties are preserved. For homogeneous toric ideals whose bouquet matrix is a monomial curve in $\mathbb{A}^3$, we give necessary and sufficient conditions for when the minimal Markov bases are distance-reducing.

Let $G$ be a cubic graph admitting a $2$-factor consisting of exactly two odd circuits, and let the complementary $1$-factor contain precisely three spokes (along with an arbitrary number of chords). We show that four perfect matchings can cover $G$. As a consequence, $G$ fulfils the 7/5-Conjecture of Alon and Tarsi.

We propose a local transformation on bicolored graphs, which we call local homophily, inspired by adaptive networks and based on majority dynamics and homophily. In this transformation, a vertex updates its color to match the majority of its neighbors, while neighbors of the same color become connected and neighbors of the opposite color become disconnected. We show how to simulate Boolean circuits using local homophily and establish that determining whether a given pair of vertices becomes connected under iterative applications of local homophily is $\mathbf{P}$-complete under logspace reductions.

The performance of the Quantum Approximate Optimization Algorithm (QAOA) on noisy intermediate-scale quantum (NISQ) devices is strongly limited by sparse qubit connectivity. When interactions required by QAOA Hamiltonians are not aligned to the hardware topology, transpilation introduces SWAP gates, increasing circuit depth and noise. We propose a SWAP-free QAOA framework based on modifying the cost Hamiltonian so that it can be implemented natively on the hardware. We formulate this as a mixed-integer semidefinite program (MISDP) that selects a hardware-compatible approximation of the original cost matrix and optimizes the allocation of logical variables to physical qubits. We prove that the associated decision problem is NP-complete and derive theoretical guarantees relating the MISDP objective to the loss in the original optimization problem through the Lovász number of the hardware graph. Since solving MISDPs is practical only for small instances, we introduce heuristics based on spectral properties of the problem matrix and hardware graph. Our experiments on a cardinality-constrained quadratic optimization model for index tracking show competitive performance against a baseline representing ideal QAOA under SWAP-induced noise. These results indicate that, on sparse NISQ architectures, a hardware-aware approximation of the objective may be more effective than an exact but heavily transpiled Hamiltonian implementation.

Many combinatorial optimization problems admit a maximin fairness variant, where the aim is to find a distribution over possible solutions which maximizes an expected worst-case outcome. However, the support for an optimal distribution may be exponential, which can be intractable to represent in the worst case. To this end, we propose a quantum based approach to solving distribution optimization problems. Expanding on work analyzing the Dynamical Lie Algebras of the Quantum Approximate Optimization Algorithm (QAOA), we show that with a finite number of layers, a QAOA ansatz can be constructed to capture any distribution over bitstrings. We show that the resulting circuit is able to effectively solve the Fair Cut Cover, a fair interpretation of the classical Fractional Cut Cover Problem. In addition, we show that our algorithm is provably better than classical approximations on certain graph structures and empirically outperforms these classical algorithms on tested instances.

We consider Boolean networks with interaction graphs partitioned into strongly connected components, which we call strong modules. This type of network decomposition has been considered in the literature, primarily from the perspective of attractor detection algorithms. In this paper, we aim to provide an algebraic basis for this line of research in the case of asynchronous Boolean networks. We prove that the asynchronous attractors of a network can be described as a dependent sum construction: as products of attractors of its controlled strong modules. We then show that a representation of all attractors can be computed in polynomial time under two conditions: the strong modules are small, and either the network is sparse or its defining functions have small size circuits (in particular when they are nested canalizing). We illustrate these results on a published Boolean model.