arXiv++ Combinatorics

We study the m-Eternal Domination problem, which is the following two-player game between a defender and an attacker on a graph: initially, the defender positions k guards on vertices of the graph; the game then proceeds in turns between the defender and the attacker, with the attacker selecting a vertex and the defender responding to the attack by moving a guard to the attacked vertex. The defender may move more than one guard on their turn, but guards can only move to neighboring vertices. The defender wins a game on a graph G with k guards if the defender has a strategy such that at every point of the game the vertices occupied by guards form a dominating set of G and the attacker wins otherwise. The m-eternal domination number of a graph G is the smallest value of k for which (G,k) is a defender win. We show that m-Eternal Domination is NP-hard, as well as some of its variants, even on special classes of graphs. We also show structural results for the Domination and m-Eternal Domination problems in the context of four types of infinite regular grids: square, octagonal, hexagonal, and triangular, establishing tight bounds.

Abstract notions of convexity over the vertices of a graph, and corresponding notions of halfspaces, have recently gained attention from the machine learning community. In this work we study monophonic halfspaces, a notion of graph halfspaces defined through closure under induced paths. Our main result is a $2$-satisfiability based decomposition theorem, which allows one to represent monophonic halfspaces as a disjoint union of certain vertex subsets. Using this decomposition, we achieve efficient and (nearly) optimal algorithms for various learning problems, such as teaching, active, and online learning. Most notably, we obtain a polynomial-time algorithm for empirical risk minimization. Independently of the decomposition theorem, we obtain an efficient, stable, and proper sample compression scheme. This makes monophonic halfspaces efficiently learnable with proper learners and linear error rate $1/\varepsilon$ in the realizable PAC setting. Our results answer open questions from the literature, and show a stark contrast with geodesic halfspaces, for which most of the said learning problems are NP-hard.

For $0\leq ρ\leq 1$ and a coloured graph $G$, a vertex $v$ is $ρ$-happy if at least $ρ\mathrm{deg}(v)$ of its neighbours have the same colour as $v$. Soft happy colouring of a partially coloured graph $G$ is the problem of finding a vertex colouring $σ$ that preserves the precolouring and has the maximum number of $ρ$-happy vertices. It is already known that this problem is NP-hard and directly relates to the community structure of the graphs; under a certain condition on the proportion of happiness $ρ$ and for graphs with community structures, the induced colouring by communities can make all the vertices $ρ$-happy. We show that when $0\leq ρ_1<ρ_2\leq 1$, a complete $ρ_2$-happy colouring has a higher accuracy of community detection than a complete $ρ_1$-happy colouring. Moreover, when $ρ$ is greater than a threshold, it is unlikely for an algorithm to find a complete $ρ$-happy colouring with colour classes of almost equal sizes. Three local search algorithms for soft happy colouring are proposed, and their performances are compared with one another and other known algorithms. Among them, the linear-time local search is shown to be not only very fast, but also a reliable algorithm that can dramatically improve the number of $ρ$-happy vertices.

A graph $G$ is perfectly divisible if every induced subgraph $H$ of $G$ contains a set $X$ of vertices such that $X$ meets all largest cliques of $H$, and $X$ induces a perfect graph. The chromatic number of a perfectly divisible graph $G$ is bounded by $ω^2$ where $ω$ denotes the number of vertices in a largest clique of $G$. A graph $G$ is minimally non-perfectly divisible if $G$ is not perfectly divisible but each of its proper induced subgraph is. A set $C$ of vertices of $G$ is a clique cutset if $C$ induces a clique in $G$, and $G-C$ is disconnected. We prove that a $P_5$-free minimally non-perfectly divisible graph cannot contain a clique cutset. This result allows us to re-establish several theorems on the perfect divisibility of some classes of $P_5$-free graphs. We will show that recognizing perfectly divisible graphs is NP-hard.

While the problem of determining the representation number of an arbitrary word-representable graph is NP-hard, this problem is open even for bipartite graphs. The representation numbers are known for certain bipartite graphs including all the graphs with at most nine vertices. For bipartite graphs with partite sets of sizes $m$ and $n$, Glen et al. conjectured that the representation number is at most $\lceil \frac{m+n}{4}\rceil$, where $m+n \ge 9$. In this paper, we show that every bipartite graph is $\left( 1+ \lceil \frac{m}{2} \rceil \right)$-representable, where $m$ is the size of its smallest partite set. Furthermore, if $m$ is odd then we prove that the bipartite graphs are $\lceil \frac{m}{2} \rceil $-representable. Accordingly, we establish that the conjecture by Glen et al. holds good for all bipartite graphs leaving the bipartite graphs whose partite sets are of equal and even size. In case of the bipartite graphs with partite sets of equal and even size, we prove the conjecture for certain subclasses using the neighborhood inclusion graph approach.

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.

A graph G(V, E) is word-representable if there exists a word w over V such that distinct letters x and y alternate in w iff $xy \in E$. We introduce p-complete squares and p-complete square-free word-representable graphs. A word is p-complete square-free if no induced subword over any subset of letters contains a square XX with $|X| \ge p$. A graph is p-complete square-free if it admits such a representation. We define p-complete square-free uniform word-representations and study their properties. We show that any graph admitting such a representation forbids Kp as an induced subgraph and that the recognition problem is NP-hard for arbitrary p. For p=1 and 2, we give complete characterisations. We prove that every $K_p$-free circle graph admits a p-complete square-free uniform representation and that any 3-complete square-free uniform word-representable graph has representation number at most three. We present a constructive method for generating new examples for p=3.

Given a graph $G$, a subset $M\subseteq V(G)$ is a mutual-visibility (MV) set if for every $u,v\in M$, there exists a $u,v$-geodesic whose internal vertices are not in $M$. We investigate proper vertex colorings of graphs whose color classes are mutual-visibility sets. The main concepts that arise in this investigation are independent mutual-visibility (IMV) sets and vertex partitions into these sets (IMV colorings). The IMV number $μ_{i}$ and the IMV chromatic number $χ_{μ_{i}}$ are defined as maximum and minimum cardinality taken over all IMV sets and IMV colorings, respectively. Along the way, we also continue with the study of MV chromatic number $χ_μ$ (as the smallest number of sets in a vertex partition into MV sets), which was initiated in an earlier paper. We establish a close connection between the (I)MV chromatic numbers of subdivisions of complete graphs and Ramsey numbers $R(4^k;2)$. From the computational point of view, we prove that the problems of computing $χ_{μ_{i}}$ and $μ_{i}$ are NP-complete, and that it is NP-hard to decide whether a graph $G$ satisfies $\imv(G)=α(G)$ where $α(G)$ is the independence number of $G$. Several tight bounds on $χ_{μ_{i}}$, $χ_μ$ and $μ_{i}$ are given. Exact values/formulas for these parameters in some classical families of graphs are proved. In particular, we prove that $χ_{μ_{i}}(T)=χ_μ(T)$ holds for any tree $T$ of order at least $3$, and determine their exact formulas in the case of lexicographic product graphs. Finally, we give tight bounds on the (I)MV chromatic numbers for the Cartesian and strong product graphs, which lead to exact values in some important families of product graphs.

Flows and colorings are disparate concepts in graph algorithms -- the former is tractable while the latter is intractable. Tutte introduced the concept of nowhere-zero flows to unify these two concepts. Jaeger showed that nowhere-zero flows are equivalent to cut-balanced orientations. Motivated by connections between nowhere-zero flows, cut-balanced orientations, Nash-Williams' well-balanced orientations, and postman problems, we study optimization versions of nowhere-zero flows and cut-balanced orientations. Given a bidirected graph with asymmetric costs on two orientations of each edge, we study the min cost nowhere-zero $k$-flow problem and min cost $k$-cut-balanced orientation problem. We show that both problems are NP-hard to approximate within any finite factor. Given the strong inapproximability result, we design bicriteria approximations for both problems: we obtain a $(6,6)$-approximation to the min cost nowhere-zero $k$-flow and a $(k,6)$-approximation to the min cost $k$-cut-balanced orientation. For the case of symmetric costs (where the costs of both orientations are the same for every edge), we show that the nowhere-zero $k$-flow problem remains NP-hard and admits a $3$-approximation.

An elimination tree of a connected graph $G$ is a rooted tree on the vertices of $G$ obtained by choosing a root $v$ and recursing on the connected components of $G-v$ to obtain the subtrees of $v$. The graph associahedron of $G$ is a polytope whose vertices correspond to elimination trees of $G$ and whose edges correspond to tree rotations, a natural operation between elimination trees. These objects generalize associahedra, which correspond to the case where $G$ is a path. Ito et al. [ICALP 2023] recently proved that the problem of computing distances on graph associahedra is NP-hard. In this paper we prove that the problem, for a general graph $G$, is fixed-parameter tractable parameterized by the distance $k$. Prior to our work, only the case where $G$ is a path was known to be fixed-parameter tractable. To prove our result, we use a novel approach based on a marking scheme that restricts the search to a set of vertices whose size is bounded by a (large) function of $k$.

We study the complexity of a class of promise graph homomorphism problems. For a fixed graph H, the H-colouring problem is to decide whether a given graph has a homomorphism to H. By a result of Hell and Nešetřil, this problem is NP-hard for any non-bipartite loop-less graph H. Brakensiek and Guruswami [SODA 2018] conjectured the hardness extends to promise graph homomorphism problems as follows: fix a pair of non-bipartite loop-less graphs G, H such that there is a homomorphism from G to H, it is NP-hard to distinguish between graphs that are G-colourable and those that are not H-colourable. We confirm this conjecture in the cases when both G and H are 4-colourable. This is a common generalisation of previous results of Khanna, Linial, and Safra [Comb. 20(3): 393-415 (2000)] and of Krokhin and Opršal [FOCS 2019]. The result is obtained by combining the algebraic approach to promise constraint satisfaction with methods of topological combinatorics and equivariant obstruction theory.

Eternal vertex cover is the following two-player game between a defender and an attacker on a graph. Initially, the defender positions k guards on k vertices of the graph; the game then proceeds in turns between the defender and the attacker, with the attacker selecting an edge and the defender responding to the attack by moving some of the guards along the edges, including the attacked one. The defender wins a game on a graph G with k guards if they have a strategy such that, in every round of the game, the vertices occupied by the guards form a vertex cover of G, and the attacker wins otherwise. The eternal vertex cover number of a graph G is the smallest number k of guards allowing the defender to win and Eternal Vertex Cover is the problem of computing the eternal vertex cover number of the given graph. We study this problem when restricted to the well-known class of series-parallel graphs. In particular, we prove that Eternal Vertex Cover can be solved in linear time when restricted to melon graphs, a proper subclass of series-parallel graphs. Moreover, we also conjecture that this problem is NP-hard on series-parallel graphs.

Parameterized Sequential Decision Making (Para-SDM) framework models a wide array of network design applications spanning supply-chain, transportation, and sensor networks. These problems entail sequential multi-stage optimization characterized by states, control actions, and cost functions dependent on designable parameters. The challenge is to determine both the sequential decision policy and parameters simultaneously to minimize cumulative stagewise costs. Many Para-SDM problems are NP-hard and often necessitate time-varying policies. Existing algorithms tackling finite-horizon time-varying Para-SDM problems struggle with scalability when faced with a large number of states. Conversely, the sole algorithm addressing infinite-horizon Para-SDM assumes time (stage)-invariance, yielding stationary policies. However, this approach proves scalable for time-invariant problems by leveraging deep neural networks to learn optimal stage-invariant state-action value functions, enabling handling of large-scale scenarios. This article proposes a novel approach that reinterprets finite-horizon, time-varying Para-SDM problems as equivalent time-invariant problems through topography lifting. Our method achieves nearly identical results to the time-varying solution while exhibiting improved performance times in various simulations, notably in the small cell network problem. This fresh perspective on Para-SDM problems expands the scope of addressable issues and holds promise for future scalability through the integration of learning methods.

A framework (a straight-line embedding of a graph into a normed space allowing edges to cross) is globally rigid if any other framework with the same edge lengths with respect to the chosen norm is an isometric copy. We investigate global rigidity in polyhedral normed spaces: normed spaces where the unit ball is a polytope. We first provide a deterministic algorithm for checking whether or not a framework in a polyhedral normed space is globally rigid. After showing that determining if a framework is globally rigid is NP-Hard, we then provide necessary conditions for global rigidity for generic frameworks. We obtain stronger results for generic frameworks in $\ell_\infty^d$ (the vector space $\mathbb{R}^d$ equipped with the $\ell_\infty$ metric) including an exact characterisation of global rigidity when $d=2$, and an easily-computable sufficient condition for global rigidity using edge colourings. Our 2-dimensional characterisation also has a surprising consequence: Hendrickson's global rigidity condition fails for generic frameworks in $\ell_\infty^2$.

We develop a structural approach to simultaneous embeddability in temporal sequences of graphs, inspired by graph minor theory. Our main result is a classification theorem for 2-connected temporal sequences: we identify five obstruction classes and show that every 2-connected temporal sequence is either simultaneously embeddable or admits a sequence of improvements leading to an obstruction. This structural insight leads to a polynomial-time algorithm for deciding the simultaneous embeddability of 2-connected temporal sequences. The restriction to 2-connected sequences is necessary, as the problem is NP-hard for connected graphs, while trivial for 3-connected graphs. As a consequence, our framework also resolves the rooted-tree SEFE problem, a natural extension of the well-studied sunflower SEFE. More broadly, our results demonstrate the applicability of graph minor techniques to evolving graph structures and provide a foundation for future algorithmic and structural investigations in temporal graph theory.

Independently posed by Behzad and Vizing, the Total Coloring Conjecture asserts that the total chromatic number of a simple connected graph $G$ is either $Δ(G)+1$ or $Δ(G)+2$, where $Δ(G)$ is the largest degree of any vertex of $G$. To decide whether a cubic graph $G$ has total chromatic number $Δ(G)+1$, even for bipartite cubic graphs, is NP-hard. The resulting problems and research persist even for total colorings that are equitable, namely with the cardinalities of the color classes differing at most by 1. Williams and Holroyd gave a new condition to solve total coloring problems via the introduction of semi-total colorings. We focus on how to obtain equitable total colorings of symmetric cubic graphs and cage graphs by means of a variation of Kempe'a 1879 graph-coloring algorithm. Such variation takes semi-total colorings to equitable ones.

We introduce achievement positional games, a convention for positional games which encompasses the Maker-Maker and Maker-Breaker conventions. We consider two hypergraphs, one red and one blue, on the same vertex set. Two players, Left and Right, take turns picking a previously unpicked vertex. Whoever first fills an edge of their color, blue for Left or red for Right, wins the game (draws are possible). We establish general properties of such games. In particular, we show that a lot of principles which hold for Maker-Maker games generalize to achievement positional games. We also study the algorithmic complexity of deciding whether Left has a winning strategy as first player when all blue edges have size at mot $p$ and all red edges have size at most $q$. This problem is in P for $p,q \leq 2$, but it is NP-hard for $p \geq 3$ and $q=2$, coNP-complete for $p=2$ and $q \geq 3$, and PSPACE-complete for $p,q \geq 3$. A consequence of this last result is that, in the Maker-Maker convention, deciding whether the first player has a winning strategy on a hypergraph of rank 4 after one round of (non-optimal) play is PSPACE-complete.

The \emph{interestingness score} of a directed path $Π= e_1, e_2, e_3, \dots, e_\ell$ in an edge-weighted directed graph $G$ is defined as $\texttt{score}(Π) := \sum_{i=1}^\ell w(e_i) \cdot \log{(i+1)}$, where $w(e_i)$ is the weight of the edge $e_i$. We consider two optimization problems that arise in the analysis of Mapper graphs, which is a powerful tool in topological data analysis. In the IP problem, the objective is to find a collection $\mathcal{P}$ of edge-disjoint paths in $G$ with the maximum total interestingness score. %; that is, two raised to the power of the sum of the weights of the paths in $\mathcal{P}$. For $k \in \mathbb{N}$, the $k$-IP problem is a variant of the IP problem with the extra constraint that each path in $\mathcal{P}$ must have exactly $k$ edges. Kalyanaraman, Kamruzzaman, and Krishnamoorthy (Journal of Computational Geometry, 2019) claim that both IP and $k$-IP (for $k \geq 3$) are NP-complete. We point out some inaccuracies in their proofs. Furthermore, we show that both problems are NP-hard in directed acyclic graphs.

Inspired by a common technique for shuffling a deck of cards on a table without riffling, we formalize the pile shuffle and investigate its capabilities as a sorting device. Our study is novel in that we consider pile shuffle in three variations: (1) using queue-like piles, (2) using stack-like piles, and (3) using a heterogeneous mixture of those two pile types, either given or decided during the shuffle by the dealer herself. We first characterize the sortable permutations of one or more sequential rounds of pile shuffle, under constraints on the number and types of piles to be used, and then derive formulas to obtain a sorting shuffle of a given permutation on the minimum number of piles, efficiently, whenever feasible. In the process, we present an interesting mathematical interpretation of any multi-round shuffle on fixed-type piles, as a single-round shuffle on a generally larger number of fixed-type ``virtual piles''; we confirm that repetition augments the power of pile shuffle exponentially, in that $m$ piles over $T$ rounds enjoys the capacity of $m^T$ piles in a single round. Finally, we motivate a forthcoming companion paper in which we prove that dealer choice -- where the dealer is allowed to choose the types of the piles during the shuffle -- makes deciding feasibility of sort NP-Hard for some variants of multi-round pile shuffle; proof is by a novel reduction from Boolean satisfiability (SAT).

In a digraph $D$, an arc $e=(x,y) $ in $D$ is considered transitive if there is a path from $x$ to $y$ in $D- e$. A digraph is transitive-free if it does not contain any transitive arc. In the Transitive-free Vertex Deletion (TVD) problem, the goal is to find at most $k$ vertices $S$ such that $D-S$ has no transitive arcs. In our work, we study a more general version of the TVD problem, denoted by $\ell$-Relaxed Transitive-free Vertex Deletion ($\ell$-RTVD), where we look for at most $k$ vertices $S$ such that $D-S$ has no more than $\ell$ transitive arcs. We explore $\ell$-RTVD on various well-known graph classes of digraphs such as directed acyclic graphs (DAGs), planar DAGs, $α$-bounded digraphs, tournaments, and their multiple generalizations such as in-tournaments, out-tournaments, local tournaments, acyclic local tournaments, and obtain the following results. Although the problem admits polynomial-time algorithms in tournaments, $α$-bounded digraphs, and acyclic local tournaments for fixed values of $\ell$, it remains NP-hard even in planar DAGs with maximum degree 6. In the parameterized realm, for $\ell$-RTVD on in-tournaments and out-tournaments, we obtain polynomial kernels parameterized by $k+\ell$ for bounded independence number. But the problem remains fixed-parameter intractable on DAGs when parameterized by $k$.