paper-with-me

홈 › Papers

SATBench: Benchmarking LLMs' Logical Reasoning via Automated Puzzle Generation from SAT Formulas

2025-05-20 · Anjiang Wei, Yuheng Wu, Yingjia Wan, Tarun Suresh, Huanmi Tan, Zhanke Zhou, Sanmi Koyejo, Ke Wang, Alex Aiken

We introduce SATBench, a benchmark for evaluating the logical reasoning capabilities of large language models (LLMs) through logical puzzles derived from Boolean satisfiability (SAT) problems. Unlike prior work that focuses on inference rule-based reasoning, which often involves deducing conclusions from a set of premises, our approach leverages the search-based nature of SAT problems, where the objective is to find a solution that fulfills a specified set of logical constraints. Each instance in SATBench is generated from a SAT formula, then translated into a story context and conditions using LLMs. The generation process is fully automated and allows for adjustable difficulty by varying the number of clauses. All 2100 puzzles are validated through both LLM-assisted and solver-based consistency checks, with human validation on a subset. Experimental results show that even the strongest model, o4-mini, achieves only 65.0% accuracy on hard UNSAT problems, close to the random baseline of 50%. SATBench exposes fundamental limitations in the search-based logical reasoning abilities of current LLMs and provides a scalable testbed for future research in logical reasoning.

📄 PDF Abstract BibTeX arXiv:2505.14615

Code (0)

등록된 구현이 없습니다.

Tasks

BenchmarkingLogical Reasoning

Methods 이 논문이 사용한 방법론

SET Dynamic Sparse Training method where weight mask is updated randomly periodically

Similar Papers 제목 키워드 기반

G4SATBench: Benchmarking and Advancing SAT Solving with Graph Neural Networks

2023-09-29 · Zhaoyu Li, Jinpei Guo, Xujie Si

Graph neural networks (GNNs) have recently emerged as a promising approach for solving the Boolean Satisfiability Problem (SAT), offering potential alternatives to traditional backtracking or local search SAT solvers. Ho…

Benchmarking

QMFOL: Benchmarking Large Language Model Reasoning via Quantifiable Monadic First-Order Logic Test Case Generation

2026-06-18 · Xinyi Zheng, Ling Shi, Tianlong Yu, Yongxin Zhao 외 arxiv

Large Language Models (LLMs) have made significant progress in reasoning, particularly in deductive reasoning, which is crucial for high-stakes decision-making. As models improve, evaluation benchmarks should evolve to k…

WildScore: Benchmarking MLLMs in-the-Wild Symbolic Music Reasoning

2025-09-05 · Gagan Mundada, Yash Vishe, Amit Namburi, Xin Xu 외 arxiv

Recent advances in Multimodal Large Language Models (MLLMs) have demonstrated impressive capabilities across various vision-language tasks. However, their reasoning abilities in the multimodal symbolic music domain remai…

Question Answering

From Knowledge to Reasoning: Evaluating LLMs for Ionic Liquids Research in Chemical and Biological Engineering

2025-05-11 · Gaurab Sarkar, Sougata Saha

Although Large Language Models (LLMs) have achieved remarkable performance in diverse general knowledge and reasoning tasks, their utility in the scientific domain of Chemical and Biological Engineering (CBE) is unclear.…

BenchmarkingGeneral Knowledge

Automated Marine Biofouling Assessment: Benchmarking Computer Vision and Multimodal LLMs on the Level of Fouling Scale

2026-01-28 · Brayden Hamilton, Tim Cashmore, Peter Driscoll, Trevor Gee 외 arxiv

Marine biofouling on vessel hulls poses major ecological, economic, and biosecurity risks. Traditional survey methods rely on diver inspections, which are hazardous and limited in scalability. This work investigates auto…