https://arxiv.org/api/F3mxro6HNLlGAGc6cIgVQYpnnh8 2026-06-21T09:00:26Z 12967 585 15 http://arxiv.org/abs/2603.17075v1 2026-03-17T19:04:37Z

Motivated by auto-proof generation and Valiant's VP vs. VNP conjecture, we study the problem of discovering efficient arithmetic circuits to compute polynomials, using addition and multiplication gates. We formulate this problem as a single-player game, where an RL agent attempts to build the circuit within a fixed number of operations. We implement an AlphaZero-style training loop and compare two approaches: Proximal Policy Optimization with Monte Carlo Tree Search (PPO+MCTS) and Soft Actor-Critic (SAC). SAC achieves the highest success rates on two-variable targets, while PPO+MCTS scales to three variables and demonstrates steady improvement on harder instances. These results suggest that polynomial circuit synthesis is a compact, verifiable setting for studying self-improving search policies.

2026-03-17T19:04:37Z ICLR 2026 Workshop on AI with Recursive Self-Improvement Weikun K. Zhang Rohan Pandey Bhaumik Mehta Kaijie Jin Naomi Morato Archit Ganapule Michael Ruofan Zeng Jarod Alper http://arxiv.org/abs/2601.09345v2 2026-03-17T10:04:04Z

Wataridori is a pencil puzzle that involves drawing paths in a rectangular grid to connect circles into pairs while satisfying several constraints. In this paper, we prove that deciding whether a given Wataridori puzzle has a solution is NP-complete via a reduction from Numberlink, another pencil puzzle that has previously been proved NP-complete.

2026-01-14T10:21:21Z Suthee Ruangwises http://arxiv.org/abs/2506.01645v2 2026-03-17T09:17:00Z

For fixed sets $σ, ρ$ of non-negative integers, the $(σ, ρ)$-domination framework introduced by Telle [Nord. J. Comput. 1994] captures many classical graph problems. For a graph $G$, a $(σ,ρ)$-set is a set $S$ of vertices such that for every $v\in V(G)$, we have (1) if $v \in S$, then $|N(v) \cap S| \in σ$, and (2) if $v \notin S$, then $|N(v) \cap S| \in ρ$. We initiate the study of a natural partial variant $(σ,ρ)$-MinParDomSet of the problem, in which the constraints given by $σ, ρ$ need not be fulfilled for all vertices, but we want to find a set of size at most $k$ that maximizes the number of vertices that are satisfied in the sense that they satisfy (1) or (2) above. Our goal is to understand whether $(σ,ρ)$-MinParDomSet can be solved in the same running time as the nonpartial version, or whether it is strictly harder. Formally, we consider nonempty finite or simple cofinite sets $σ$ and $ρ$ (simple cofinite sets are of the form $\mathbb{Z}_{\geq c}$), and we try to determine the smallest constant $c_{σ,ρ}$ such that there is a $c_{σ,ρ}^{tw}\cdot n^{O(1)}$ time algorithm for the problem if a tree decomposition of width $tw$ is given. We obtain matching upper and lower bounds on $c_{σ,ρ}$ for every such fixed $σ$ and $ρ$ under the Primal Pathwidth Strong Exponential Time Hypothesis, and establish whether the partial problem is harder than the nonpartial variant. For some sets $σ$ and $ρ$, the more general $(σ,ρ)$-MinParDomSet has the same complexity as the nonpartial special case (e.g., for Dominating Set), while for other choices, the partial version is significantly harder (e.g., for Perfect Code).

2025-06-02T13:17:45Z Abstract shortened; adds a lower bound for the nonpartial problem and improves exposition Jakob Greilhuber Dániel Marx http://arxiv.org/abs/2603.16156v1 2026-03-17T06:20:28Z

We prove that there exists a deterministic configuration of Conflict Driven Clause Learning (CDCL) SAT solvers using a variant of the VSIDS branching heuristic that solves instances of the Ordering Principle (OP) CNF formulas in time polynomial in n, where n is the number of variables in such formulas. Since tree-like resolution is known to have an exponential lower bound for proof size for OP formulas, it follows that CDCL under this configuration has an exponential separation with any solver that is polynomially equivalent to tree-like resolution and therefore any configuration of DPLL SAT solvers.

2026-03-17T06:20:28Z Sahil Samar Marc Vinyals Vijay Ganesh http://arxiv.org/abs/2603.16029v1 2026-03-17T00:29:54Z

We develop a perfectly distributable quantum-classical streaming algorithm that processes signed edges to efficiently estimate the counts of triangles of diverse signed configurations in the single pass edge stream. Our approach introduces a quantum sketch register for processing the signed edge stream, together with measurement operators for query-pair calls in the quantum estimator, while a complementary classical estimator accounts for triangles not captured by the quantum procedure. This hybrid design yields a polynomial space advantage over purely classical approaches, extending known results from unsigned edge stream data to the signed setting. We quantify the lack of balance on random signed graph instances, showcasing how the classical and hybrid algorithms estimate balance in practice.

2026-03-17T00:29:54Z 35 pages Steven Kordonowy Bibhas Adhikari Hannes Leipold http://arxiv.org/abs/2603.15856v1 2026-03-16T19:38:57Z

Fix a finite field $\mathbb F_q$ and let $A\in \mathbb F_q^{n\times n}$ be a uniformly random $n\times n$ matrix over $\mathbb F_q$. The asymptotic distribution of the determinant $\det(A)$ is well-understood, but the asymptotic distribution of the permanent $\operatorname{per}(A)$ is still something of a mystery. In this paper we make a first step in this direction, proving that $\operatorname{per}(A)$ is significantly more uniform than $\det(A)$.

2026-03-16T19:38:57Z Zach Hunter Matthew Kwan Lisa Sauermann http://arxiv.org/abs/2603.15565v1 2026-03-16T17:26:33Z

We prove two new upper bounds for depth-2 linear circuits computing the $N$th disjointness matrix $D^{\otimes N}$. First, we obtain a circuit of size $O\big(2^{1.24485N}\big)$ over $\{0,1\}$. Second, we obtain a circuit of degree $O\big(2^{0.3199N}\big)$ over $\{0,\pm 1\}$. These improve the previous bounds of Alman and Li, namely size $O\big(2^{1.249424N}\big)$ and degree $O\big(2^{N/3}\big)$. Our starting point is the rebalancing framework developed in a line of works by Jukna and Sergeev, Alman, Sergeev, and Alman-Guan-Padaki, culminating in Alman and Li. We sharpen that framework in two ways. First, we replace the earlier "wild" rebalancing process by a tame, discretized process whose geometric-average behavior is governed by the quenched top Lyapunov exponent of a random matrix product. This allows us to invoke the convex-optimization upper bound of Gharavi and Anantharam. Second, for the degree bound we work explicitly with a cost landscape on the $(p,q)$-plane and show that different circuit families are dominant on different regions, so that the global maximum remains below $0.3199$.

2026-03-16T17:26:33Z 11 pages Lixi Ye http://arxiv.org/abs/2603.11996v2 2026-03-16T13:28:27Z

Submodular maximization constitutes a prominent research topic in combinatorial optimization and theoretical computer science, with extensive applications across diverse domains. While substantial advancements have been achieved in approximation algorithms for submodular maximization, the majority of algorithms yielding high approximation guarantees are randomized. In this work, we investigate deterministic approximation algorithms for maximizing non-monotone submodular functions subject to matroid and knapsack constraints. For the two distinct constraint settings, we propose novel deterministic algorithms grounded in an extended multilinear extension framework. Under matroid constraints, our algorithm achieves an approximation ratio of $(0.385 - ε)$, whereas for knapsack constraints, the proposed algorithm attains an approximation ratio of $(0.367 -ε)$. Both algorithms run in $\mathrm{poly}(n)$ query complexity, where $n$ is the size of the ground set, and improve upon the state-of-the-art deterministic approximation ratios of $(0.367 - ε)$ for matroid constraints and $0.25$ for knapsack constraints.

2026-03-12T14:46:08Z Shengminjie Chen Yiwei Gao Kaifeng Lin Xiaoming Sun Jialin Zhang http://arxiv.org/abs/2502.06740v4 2026-03-16T10:06:21Z

The central open question of algebraic complexity is whether VP is unequal to VNP, which is saying that the permanent cannot be represented by families of polynomial-size algebraic circuits. For symmetric algebraic circuits, this has been confirmed by Dawar and Wilsenach (2020, 2025) who showed exponential lower bounds on the size of symmetric circuits for the permanent. In this work, we set out to develop a more general symmetric algebraic complexity theory. Our main result is that a family of symmetric polynomials admits small symmetric circuits if and only if they can be written as a linear combination of homomorphism counting polynomials of graphs of bounded treewidth. We also establish a relationship between the symmetric complexity of subgraph counting polynomials and the vertex cover number of the pattern graph. As a concrete example, we examine the symmetric complexity of immanant families (a generalisation of the determinant and permanent) and show that a known conditional dichotomy due to Curticapean (2021) holds unconditionally in the symmetric setting.

2025-02-10T18:08:48Z minor updates to the ITCS 2026 version Anuj Dawar Benedikt Pago Tim Seppelt http://arxiv.org/abs/2602.01042v2 2026-03-16T06:56:50Z

Given an $n$-bit Boolean function with a complexity measure (such as block sensitivity, query complexity, etc.) $M(f) = k$, the hardness condensation question asks whether $f$ can be restricted to $O(k)$ variables such that the complexity measure is $Ω(k)$? In this work, we study the condensability of block sensitivity, certificate complexity, AND (and OR) query complexity and Fourier sparsity. We show that block sensitivity does not condense under restrictions, unlike sensitivity: there exists a Boolean function $f$ with query complexity $k$ such that any restriction of $f$ to $O(k)$ variables has block sensitivity $O(k^{\frac{2}{3}})$. This answers an open question in Göös, Newman, Riazanov, and Sokolov (2024) in the negative. The same function yields an analogous incondensable result for certificate complexity. We further show that $\mathsf{AND}$(and $\mathsf{OR}$) decision trees are also incondensable.

2026-02-01T06:03:31Z v2: Removed a proof that was a minor modification of a result from Hart, Hosseini, and Iosevich (Condensing Hardness in Boolean Functions) Sai Soumya Nalli Karthikeya Polisetty Jayalal Sarma http://arxiv.org/abs/2603.14754v1 2026-03-16T02:39:08Z

We investigate the fine-grained complexity of dynamically maintaining the result of fixed self-join free conjunctive queries under single-tuple updates. Prior work shows that free-connex queries can be maintained in update time $O(|D|^δ)$ for some $δ\in [0.5, 1]$, where $|D|$ is the size of the current database. However, a gap remains between the best known upper bound of $O(|D|)$ and lower bounds of $Ω(|D|^{0.5-ε})$ for any $ε\ge 0$. We narrow this gap by introducing two structural parameters to quantify the dynamic complexity of a conjunctive query: the height $k$ and the dimension $d$. We establish new fine-grained lower bounds showing that any algorithm maintaining a query with these parameters must incur update time $Ω(|D|^{1-1/\max(k,d)-ε})$, unless widely believed conjectures fail. These yield the first super-$\sqrt{|D|}$ lower bounds for maintaining free-connex queries, and suggest the tightness of current algorithms when considering arbitrarily large $k$ and~$d$. Complementing our lower bounds, we identify a data-dependent parameter, the generalized $H$-index $h(D)$, which is upper bounded by $|D|^{1/d}$, and design an efficient algorithm for maintaining star queries, a common class of height 2 free-connex queries. The algorithm achieves an instance-specific update time $O(h(D)^{d-1})$ with linear space $O(|D|)$. This matches our parameterized lower bound and provides instance-specific performance in favorable cases.

2026-03-16T02:39:08Z Accepted in PODS 2026 Qichen Wang http://arxiv.org/abs/2603.14744v1 2026-03-16T02:30:52Z

Cardinality-constrained binary optimization is a fundamental computational primitive with broad applications in machine learning, finance, and scientific computing. In this work, we introduce a Grover-based quantum algorithm that exploits the structure of the fixed-cardinality feasible subspace under a natural promise on solution existence. For quadratic objectives, our approach achieves ${O}\left(\sqrt{\frac{\binom{n}{k}}{M}}\right)$ Grover rotations for any fixed cardinality $k$ and degeneracy of the optima $M$, yielding an exponential reduction in the number of Grover iterations compared with unstructured search over $\{0,1\}^n$. Building on this result, we develop a hybrid classical--quantum framework based on the alternating direction method of multipliers (ADMM) algorithm. The proposed framework is guaranteed to output an $ε$-approximate solution with a consistency tolerance $ε+ δ$ using at most $ {O}\left(\sqrt{\binom{n}{k}}\frac{n^{6}k^{3/2} }{ \sqrt{M}ε^2 δ}\right)$ queries to a quadratic oracle, together with ${O}\left(\frac{n^{6}k^{3/2}}{ε^2δ}\right)$ classical overhead. Overall, our method suggests a practical use of quantum resources and demonstrates an exponential improvements over existing Grover-based approaches in certain parameter regimes, thereby paving the way toward quantum advantage in constrained binary optimization.

2026-03-16T02:30:52Z 19 pages Haomu Yuan Hanqing Wu Kuan-Cheng Chen Bin Cheng Crispin H. W. Barnes http://arxiv.org/abs/2006.01256v2 2026-03-15T12:21:34Z

An open-close door gadget has two states and three tunnels that can be traversed by an agent (player, robot, etc.): the "opening" and "closing" tunnels set the gadget's state to open and closed, respectively, while the "traverse" tunnel can be traversed if and only if the door is in the open state. We prove that it is PSPACE-complete to decide whether an agent can move from one location to another through a planar system of any such door gadget, removing the traditional need for crossover gadgets and thereby simplifying past PSPACE-hardness proofs of Lemmings and Nintendo games Super Mario Bros., Legend of Zelda, and Donkey Kong Country. Even stronger, we show that any gadget in the motion-planning-through-gadgets framework can be simulated by a planar system of door gadgets: the open-close door gadget is a universal gadget. We prove that these results hold for a variety of door gadgets. In particular, the opening, closing, and traverse tunnel locations can have an arbitrary cyclic order around the door; each tunnel can be directed or undirected; and the opening tunnel can instead be an optional button (with identical entrance and exit locations). Furthermore, we show the same hardness and universality results for two simpler types of door gadgets: self-closing door gadgets and symmetric self-closing door gadgets. Again we show that any self-closing door gadget planarly simulates any gadget, and thus the reachability motion planning problem is PSPACE-complete. Then we apply this framework to prove new PSPACE-hardness results for eight different 3D Mario video games and Sokobond.

2020-06-01T20:47:51Z 36 pages, 35 figures. All cases are now proved PSPACE-complete. New universality proofs. Earlier version published at FUN 2020 MIT Gadgets Group Jeffrey Bosboom Erik D. Demaine Jenny Diomidova Dylan Hendrickson Hayashi Layers Jayson Lynch http://arxiv.org/abs/2507.23563v4 2026-03-15T08:22:50Z

In this monograph, we study complexity classes that are defined using $O(\log n)$-space bounded non-deterministic Turing machines. We prove salient results of Computational Complexity in this topic such as the Immerman-Szelepcsenyi Theorem, the Isolating Lemma, theorems of Meena Mahajan and V. Vinay on the determinant and many consequences of these very important results. The manuscript is intended to be a comprehensive textbook on the topic of The Complexity of Logarithmic Space Bounded Counting Classes.

2025-07-31T13:50:21Z Minors typos, errors in almost all the chapters have been corrected. Some figures have been redrawn. Main results in the last chapter have been checked for correctness and re-written to improve clarity and precision T. C. Vijayaraghavan http://arxiv.org/abs/2512.23680v2 2026-03-13T21:44:46Z

As the class $\mathcal T_4$ of graphs of twin-width at most 4 contains every finite subgraph of the infinite grid and every graph obtained by subdividing each edge of an $n$-vertex graph at least $2 \log n$ times, most NP-hard graph problems, like Max Independent Set, Dominating Set, Hamiltonian Cycle, remain so on $\mathcal T_4$. However, Min Coloring and k-Coloring are easy on both families because they are 2-colorable and 3-colorable, respectively. We show that Min Coloring is NP-hard on the class $\mathcal T_3$ of graphs of twin-width at most 3. This is the first hardness result on $\mathcal T_3$ for a problem that is easy on cographs (twin-width 0), on trees (whose twin-width is at most 2), and on unit circular-arc graphs (whose twin-width is at most 3). We also show that for every $k \geqslant 3$, k-Coloring is NP-hard on $\mathcal T_4$. We finally make two observations: (1) there are currently very few problems known to be in P on $\mathcal T_d$ (graphs of twin-width at most $d$) and NP-hard on $\mathcal T_{d+1}$ for some nonnegative integer $d$, and (2) unlike $\mathcal T_4$, which contains every graph as an induced minor, the class $\mathcal T_3$ excludes a fixed planar graph as an induced minor; thus it may be viewed as a special case (or potential counterexample) for conjectures about classes excluding a (planar) induced minor. These observations are accompanied by several open questions.

2025-12-29T18:36:19Z 10 pages, 2 figures Édouard Bonnet