TTT-Bench: A Benchmark for Evaluating Reasoning Ability with Simple and Novel Tic-Tac-Toe-style Games
This work addresses the underexplored ability of LRMs to reason in broader task domains beyond STEM, though it is incremental as it introduces a new benchmark without proposing a novel method for improvement.
The authors tackled the problem of evaluating reasoning abilities in large reasoning models (LRMs) by introducing TTT-Bench, a benchmark with simple Tic-Tac-Toe-style games, and found that models excelling at hard math problems often fail at these tasks, scoring on average 41% and 5% lower compared to other benchmarks.
Large reasoning models (LRMs) have demonstrated impressive reasoning capabilities across a broad range of tasks including Olympiad-level mathematical problems, indicating evidence of their complex reasoning abilities. While many reasoning benchmarks focus on the STEM domain, the ability of LRMs to reason correctly in broader task domains remains underexplored. In this work, we introduce \textbf{TTT-Bench}, a new benchmark that is designed to evaluate basic strategic, spatial, and logical reasoning abilities in LRMs through a suite of four two-player Tic-Tac-Toe-style games that humans can effortlessly solve from a young age. We propose a simple yet scalable programmatic approach for generating verifiable two-player game problems for TTT-Bench. Although these games are trivial for humans, they require reasoning about the intentions of the opponent, as well as the game board's spatial configurations, to ensure a win. We evaluate a diverse set of state-of-the-art LRMs, and \textbf{discover that the models that excel at hard math problems frequently fail at these simple reasoning games}. Further testing reveals that our evaluated reasoning models score on average $\downarrow$ 41\% \& $\downarrow$ 5\% lower on TTT-Bench compared to MATH 500 \& AIME 2024 respectively, with larger models achieving higher performance using shorter reasoning traces, where most of the models struggle on long-term strategic reasoning situations on simple and new TTT-Bench tasks.