TTT-Bench: A Benchmark for Evaluating Reasoning Ability with Simple and Novel Tic-Tac-Toe-style Games

TL;DR

This paper introduces TTT-Bench, a new benchmark using simple Tic-Tac-Toe-style games to evaluate reasoning abilities of large reasoning models, revealing they often struggle with basic strategic reasoning despite excelling at complex math problems.

Contribution

The paper presents TTT-Bench, a scalable, verifiable benchmark for assessing basic reasoning skills in LRMs through simple yet challenging Tic-Tac-Toe-style games, highlighting gaps in current models.

Findings

Models excel at complex math but fail at simple reasoning tasks.

Performance drops by 41% compared to math benchmarks.

Larger models perform better with shorter reasoning traces.

Abstract

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.