arXiv++ Combinatorics

The independence polynomial of a simple graph $G$ is given by \( I_G(z) = i_0 + i_1 z + i_2 z^2 + \cdots + i_αz^α\), where \( i_α\) denotes the size of a maximum independent set, also called the independence number of the graph. The independence polynomial has the notable feature of being essentially closed under graph composition (lexicographic product). In this paper, we determine the independence polynomials of size five. For a disconnected graph $G$, we exploit the fact that $I_G(z)$ factors as the product of the independence polynomials of the connected components of $G$. Furthermore, we classify all independence polynomials that can occur for such a disconnected graph $G$ and, by examining their component structures, we characterize the disconnected configurations that may arise.

We give a closed formula for the number $p^{S_2}(n)$ of self-avoiding polygons (SAPs) of length $n$ on the strip $S_2:=\mathbb{Z}\times\{0,1,2\}$, together with closed formulas for those subtypes of SAPs which are determined by the numbers of vertical steps in their leftmost and rightmost columns. For the subtype whose leftmost and rightmost columns each contain two vertical steps, we also derive an alternative representation as a binomial sum. Our derivation is elementary: it is purely combinatorial and geometric and avoids generating functions. Comparing the two representations yields a new geometric proof of an identity arising in Larsen's treatment \cite{L07} of a problem posed by Gessel \cite{G95}. Finally, we show that this subtype of SAPs is closely connected to the sequence A007909. More precisely, for $m\geq0$, the number of these SAPs whose leftmost and rightmost columns each contain two vertical steps and whose length equals $2m+6$ is given by the term of this sequence with index $m$, which thereby acquires a geometric interpretation alongside the compositions it enumerates.

We disprove Hoàng's conjecture that a minimally nonperfectly divisible graph cannot contain a bisimplicial vertex. Our counterexample has 93 vertices, 320 edges, clique number three, and a bisimplicial vertex of degree four. A perfect division of a graph $H$ is a partition $V(H)=A\mathbin{\dot\cup}B$ such that $H[A]$ is perfect and $ω(H[B])<ω(H)$; a graph is perfectly divisible when each of its induced subgraphs with at least one edge has such a division. The constructed graph has no perfect division, whereas every proper induced subgraph is perfectly divisible. Its construction uses a forcing mechanism based on rooted graphs: six 15-vertex rooted graphs force their roots into $A$, and a nine-vertex auxiliary graph then forces a triangle into $B$. We isolate this mechanism in a rooted composition lemma, which proves symbolically that every proper induced subgraph is perfectly divisible. All finite assertions concerning the rooted and auxiliary graphs are verified by exact exhaustive computation, with an independent implementation providing a cross-check. The same construction gives a negative answer to a prescribed-vertex problem of Hu, Xu and Zhuang.

We have previously given a factorization of a symplectic column under the action of the parity involution which enabled to explicitly have written the inverse of the reduction map in the quantum Littlewood-Richardson bijection. Watanabe has written the reduction map as a composition of several maps, among them, combinatorial $R$-matrices on single columns and a reduction map on a shorter column with rank reduced by one. We now use this approach to write the inverse of the reduction map on a symplectic column by detecting the corresponding symplectic column of rank reduced by one and thus avoiding going through several map compositions.

We show that there exist graphons that interpolate between Riemannian manifolds and weighted geometric graphs. Specifically, the graph-to-manifold approximation used in manifold learning can be regarded as the composition of a graph-to-graphon convergence and a graphon-to-manifold convergence in a certain sense. Furthermore, we establish a monotonicity inequality which reveals an implicit relationship between numerous combinatorial parameters and geometric quantities on graphons. Using this inequality, we find relations among conductance, maxcut problem, capacity, and packing radius, as well as their limiting behaviours under graph-to-graphon and graphon-to-manifold convergences; some of these relations are novel even for simple graphs and closed manifolds.

Amdeberhan, Shareshian, and Stanley recently proved that a function $\varphi$ arising in the theory of partition Eisenstein series counts the alternating permutations of $\{1,\dots,2n\}$ with a given `record' partition, and they asked whether there is a similar theory for record compositions, suggesting a role for noncommutative symmetric functions. Here we solve their open problem by showing that the number of alternating permutations of $\{1,\dots,2n\}$ with record composition $(α_1,\dots,α_\ell)$ is \[ \prod_{j=1}^{\ell}\binom{2s_j-1}{2α_j-1}E_{2α_j-1}, \] where $s_j=α_1+\dots+α_j$, $E_k$ is an Euler number, and the record composition of $w=a_1a_2\dots a_{2n}$ (so $a_1>a_2<a_3>\dotsb$) lists the factor lengths obtained by cutting $a_1a_3\dots a_{2n-1}$ before each left-to-right maximum other than the first. These numbers are the coefficients of a natural lift of the degree-$n$ sprout symmetric function with seed $\sec(\sqrt{t}\,)$ to noncommutative symmetric functions, expanded in products of noncommutative power sums of the first kind. An analogous refinement holds for every sprout sequence whose seed is given by the exponential formula. AxiomProver autonomously produced and verified the results in this paper in Lean.

We define an involution $h$ on Catalan structures through an abstract framework, prove an equidistribution theorem for four canonical statistics and present a generating function carrying these. This framework encompasses all combinatorial structures with a decomposition mirroring the first-return decomposition of Dyck paths. The fixed points of~$h$ are counted by Catalan numbers. Canonical bijections transport the equidistribution to eight well known concrete families, identifying the canonical statistics with native ones on each. In addition to its primary structure, each Catalan structure has a derived \emph{secondary structure}, and $h$~interchanges primary and secondary structure. The involution factors as $h = \rev \circ \corev \circ \rev$, where $\rev$ and $\corev$ are two simpler involutions, and the composition $M = h \circ \rev$ coincides with Donaghey's automorphism on plane trees. This yields $M^{-1} = \rev \circ M \circ \rev$ and a period theorem: Iterating the secondary structure construction produces a sequence that repeats with period equal to the order of~$M$. It is an open problem to describe $h$ and the canonical statistics explicitly on most of the more than two hundred known families of Catalan structures.

We study the image of a regular unipotent element under any finite-dimensional irreducible polynomial representations of $\mathrm{GL}_3(\mathbb{C})$. This problem is equivalent to decomposing certain compositions of irreducible representations as $\mathrm{SL}_2(\mathbb{C})$-modules. We give an explicit decomposition of this finding, its Jordan decomposition.

We study the two-color distinct-part series \(S_1(q)\), equivalently Andrews' generating function \(v_d(q)\) for strictly concave compositions, and its odd and even companions \(T_o(q)\) and \(T_e(q)\). We determine the coefficients of \(S_1(q)\) modulo \(4\) and obtain a complete criterion for the resulting Ramanujan-type progressions. For the even companion, we give a direct overpartition interpretation of its coefficients and show that two natural partition families are each counted by half of those coefficients. For the eta-normalized odd companion \(C(q)=(q;q)_\infty T_o(q)\), we prove a quintic self-similarity, derive exact vanishing relations and infinite sign changes for its coefficients, and show that \(c(n)\) can be nonzero only when \(24n+28\) is represented by \(x^2+3y^2\).

Motivated by the twin homomorphism problem for Coxeter groups and the corresponding Hecke monoids, we find a large class of its solutions originating from standard homomorphisms of Artin monoids and their compositions. These homomorphisms are expected to be injective when they are optimal and injective on generators, which generalizes the homogeneous homomorphisms and the famous Tits conjecture settled by Crisp and Paris. We classify disjoint standard homomorphisms and conjecture the complete classification when the domain is of rank two.

In arXiv:2603.16698v5 we have explicitly computed the inverse of the reduction map in the quantum Littlewood-Richardson bijection for certain patterns of symplectic columns. It turns out that some of those patterns are cell pieces to compute the inverse of the reduction map on any symplectic column. For large symplectic columns, the tools provided here can be combined with the composition of the inverses of the several maps in which the reduction map decomposes, given by Watanabe, namely, among them, combinatorial $R$-matrices and reduction maps of shorter symplectic columns.

In 2022 Aliniaeifard, Li, and van Willigenburg defined Schur functions in the algebra of symmetric functions in noncommuting variables (NCSym), answering an open question posed by Rosas and Sagan in 2004. These Schur functions are not monomial positive, since they are defined via a noncommutative analogue of the Jacobi-Trudi determinant. We introduce Hall-Littlewood functions ${\bf P}_π({\bf x};t)$ indexed by set partitions $π$ in noncommuting variables ${\bf x}=({\bf x}_1,{\bf x}_2,\ldots)$, and define Schur functions in noncommuting variables to be ${\bf s}_π({\bf x})={\bf P}_π({\bf x};0)$. We prove that the set of Hall-Littlewood functions $\{{\bf P}_π({\bf x};t)\}$ for all set partitions $π$ of $[n]$ forms a $\mathbb{Q}[t]$-basis of NCSym of homogeneous degree $n$, and that this basis is invariant under any permutation acting on set partitions. These Hall-Littlewood functions in NCSym map to classical Hall-Littlewood functions under commutation, up to a scalar factor. We also show that the Hall-Littlewood functions ${\bf P}_π({\bf x};t)$ naturally refine the lifted Hall-Littlewood functions in NCSym. Specifically, the Schur functions ${\bf s}_π({\bf x})$ are monomial positive and refine the lifted Schur function introduced by Rosas and Sagan. Moreover, we introduce a star product of two polynomials in NCSym and develop the star-multiplication rule for a lifted and a non-lifted Hall-Littlewood functions in NCSym. This rule is a noncommutative analogue of the product rule for two Hall-Littlewood functions and, in particular, of the Littlewood-Richardson rule. Finally, our approach extends to the algebra of quasisymmetric functions in noncommuting variables (NCQSym) indexed by set compositions.

A (strongly) concave composition of an integer $n$ is a sequence of positive integers that is (strictly) decreasing to a point and then (strictly) increasing thereafter, such that the sum of the entries equals $n$. The value at the low point is called the center part. The difference between the number of entries before and after the low point of the sequence is referred to as the rank of the (strongly) concave composition. The rank functions $V_d(m,n)$ and $V(m,n)$ are defined as the number of concave compositions and strongly concave compositions, respectively, of $n$ with rank $m$. By constructing the difference systems that characterize the rank generating functions, we establish monotonicity properties for the rank functions of both strongly concave compositions and concave compositions for all positive integers $n$. Moreover, we also study the monotonicity properties for the rank functions of (strongly) concave compositions with fixed center parts.

Complementarity is the case in which a human--AI interaction (HAI) outperforms the best prediction benchmark available among its members. Although this idea is central in HAI research, formal work on complementarity remains limited. Existing frameworks do not model how agents' predictions compose into workflow-sensitive multi-agent protocols. We close this gap by introducing a tree-based formalization of complementarity in multi-agent HAI. An HAI protocol is represented by an ordered agent-role configuration together with a rooted planar binary tree whose leaves are decorated by prediction vectors. A local binary composition rule is evaluated recursively along the tree, yielding a tree-relative complementarity functional relative to a pointwise-min oracle benchmark. We prove four results. First, selector-based HAIs, including self- or AI-reliance, cannot achieve complementarity regardless of task, loss, or prediction quality. Second, in regression under squared loss, complementarity is equivalent to Euclidean distance minimization from the ground-truth vector; for $N=2$, the optimal linear-pooling weight has a closed form and a residual-correction interpretation. Third, under linear local composition, every protocol tree defines a barycentric coordinate chart on the simplex of leaf weights; Tamari-cover reparameterizations of protocol trees preserve complementarity, and for $N=4$, they satisfy the pentagon identity. Fourth, in binary classification, no internal local composition can achieve complementarity under endpoint-monotone losses, including standard Bregman and many finite Bernoulli $f$-divergence losses; an analogous obstruction holds for multiclass aggregation under cross-entropy. In summary, our framework shows that complementarity is attainable in multi-agent regression, but obstructed in classification under natural conditions on local aggregation and loss functions.

This paper is motivated by a broad question about AI-assisted mathematics: can an AI system help discover and certify an explicit bijection between two infinite sequences of complicated combinatorial sets already known to be equinumerous? The challenge is to find a reversible structure explaining that equality uniformly across the sequence. We give an affirmative test case in the setting of a partition problem. Andrews and Dhar introduced two partition families $\mathcal{C}_3(n)$ and $\mathcal{D}_3(n)$, and for "nonexceptional'' $n$, they asked for a bijective proof of their equality \[ |\mathcal{C}_3(n)|=\frac{|\mathcal{D}_3(n)|}{3}. \] We prove a residue-class equidistribution theorehm for $\mathcal{D}_3(n)$ that identifies a "canonical third'' subset $\mathcal{D}_3^{(0)}(n)\subseteq \mathcal{D}_3(n)$. Answering their question, we construct a bijection \[ ι_n:\mathcal{C}_3(n)\longrightarrow \mathcal{D}_3^{(0)}(n) \] as a highly structured composition of four maps. AxiomProver autonomously produced and Lean-verified the equidistribution theorem. The bijection was found through human--AxiomProver collaboration, and the theorem was autoformalized and verified by the system.

Combinatorial counting problems pervade artificial intelligence, statistics, and discrete mathematics. Whether the task is enumerating subsets, multisets, permutations, partitions, or compositions under structural and arithmetic constraints, solving it remains a stubbornly manual exercise. Closed-form derivations are powerful but brittle, while naive encodings to propositional model counting or constraint satisfaction destroy the exchangeability that makes counting tractable in the first place. We present Cofola (COmbinatorial counting LAnguage with First-Order logic), a typed declarative language whose primitives are the combinatorial objects that recur in everyday counting questions, including sets, bags, tuples, sequences, circles, partitions, and compositions, together with natural relational and arithmetic constraints over them. A denotational semantics maps every Cofola program to a well-defined combinatorial counting problem, and a three-phase compilation pipeline (preprocessing, decomposition, and symmetry-preserving encoding) reduces this problem to a weighted first-order model counting (WFOMC) instance augmented with coefficient-extraction constraints. To stay inside known domain-liftable fragments whenever possible, the encoding groups indistinguishable entities, breaks the symmetry of unordered groupings lexicographically, and encodes sequences and circles via order axioms. On a suite of representative combinatorial counting problems, ranging from textbook math problems to multi-object scenarios that the closest prior framework cannot express, Cofola produces concise specifications and a uniform solving pipeline that is practical end-to-end.

We develop a framework for analyzing parameter symmetries in deep ReLU networks and obtain a complete characterization of the generic parameter fibers for three-layer bottleneck architectures. Our approach provides explicit semi-algebraic descriptions of these fibers and yields a polynomial time algorithm for deciding functional equivalence of two parameters. The symmetries include discrete and continuous transformations arising from layer composition, and depend on whether deeper layers hide or preserve geometric structure from preceding layers. Finally, we show that some of these symmetries induce local conservation laws along gradient flow, while others do not.

We introduce and study integral planes associated with crystallographic and non-crystallographic integral systems in real composition algebras. For an integral order $\Order$ in such an algebra we define the plane $\Order^{2}$ with quadratic form $Q(x,y)=\NN(x)+\NN(y)$, the axis shell, the balanced shell, and the corresponding unit-normalised spherical polytopes. For ten crystallographic orders we recover, in one uniform construction, the orthogonal-direct-sum root systems $2A_{1}$, $A_{2}\oplus A_{2}$, $4A_{1}$, $D_{4}\oplus D_{4}$, $16A_{1}$, and $E_{8}\oplus E_{8}$ (with classical-polytope realisations including the square, the 16-cell, the 24-cell, and the Gosset polytope $4_{21}$); for two non-crystallographic orders we obtain $H_{2}\oplus H_{2}$ (decagonal tegum) and $H_{4}\oplus H_{4}$ (600-cell tegum) over $\Z[\golden]$. We prove a rank-obstruction theorem that closes, unconditionally and by a purely Coxeter-theoretic argument, the existence question for an indecomposable rank-eight golden octonion order: no such order can exist. On the balanced shell side, we identify the genuine algebraic Hopf map $\Hopfmap_{A}(a,b)=(2a\bar b,\NN(a)-\NN(b))$ and prove that its restriction to the balanced shell is a finite principal fibration of the unit loop, valid both for the associative case and for the alternative Moufang case.

We develop the ARE method (Action-Rectification-Expansion), a structural framework for the organization of Leibniz terms in determinants through cyclic group actions and orbital decompositions. The symmetric group S_n is partitioned into (n-1)! disjoint orbits of size n under right composition by the cyclic group C_n. Each orbit admits a canonical representative and generates a family of determinant terms related by cyclic rotation. We prove explicit sign laws for orbital rotations, establish a rectification theorem transforming orbital polylines into parallel-line configurations through a single block permutation, and characterize companion orbitals through dihedral symmetries. The framework yields an exact reorganization of the Leibniz expansion preserving all n! terms while exposing hidden geometric and combinatorial structure. We further prove an impossibility theorem showing that no fixed-width direct extension of the classical Sarrus rule can capture all determinant terms for n >= 4. The method provides three equivalent visualizations: polylines, parallel rectified lines, and total-line representations. Deterministic orbital generation algorithms and computational verification against standard determinant methods are also presented. Although the approach does not reduce factorial complexity, it provides a systematic geometric and algebraic interpretation of determinant structure extending the conceptual spirit of Sarrus to arbitrary dimension.

The alternated composition of $N=2p$ differential operators $ w_j(x)\,\partial_x^p$ of strict order $p$ on the line $\mathbb{R}\ni x$ is again a differential operator of strict order $p$; its coefficient is the constant $\mathrm{const}(p)$, depending only on the arity $N$, times the Wronskian determinant of the originally taken coefficients $w_1$, $\ldots$, $w_N$. The case $p=1$ of the Lie bracket for two vector fields fixes $\mathrm{const}(1)=1$. When $p=2$, finding $\mathrm{const}(2)=2$ is easy; we obtain $\mathrm{const}(3)=90$. The problem is to know $\mathrm{const}(p\geqslant 4)$. We express the formula of $\mathrm{const}(p)$ in terms of the sum with signs over the much smaller set of 'late-growing' permutations, thus reaching the exact values $c(p=4)= 586\,656$, $c(p=5)\approx 1.9\cdot 10^{12}$, and $c(p=6)\approx 7.9\cdot 10^{21}$; the positive integer sequence $\mathrm{const}(p)$ seems to be new.