arXiv cs.CL· Ori Press, Alexander Golubev·· 12 小时前AI 评分38
用自然语言解释国际象棋与围棋：BinProp 让语言模型发现可复用棋类概念
Explaining Chess and Go in Natural Language
AI 导读
研究提出一种基于棋类引擎评估的自动指标，用于衡量语言模型对国际象棋走法的解释质量，发现当前最先进模型表现意外地差，常依赖表层棋盘观察而非可迁移的棋类概念。为此提出免训练的 BinProp 方法，让同一批模型自行发现可复用的棋类概念并用自然语言解释走法，所有被测模型的平均得分接近提升至直接提示的 3 倍。该方法无需额外调优即可在围棋、跳棋和将棋中发现有用概念。
正文 · 原文
Abstract:Modern chess engines play far beyond human ability, often choosing moves whose underlying ideas are difficult for humans to understand. We propose an automatic metric for evaluating chess explanations, grounded in chess-engine evaluations. Evaluating state-of-the-art language models using this metric, we find that they perform surprisingly poorly, often relying on surface-level board observations rather than robust, transferable chess concepts. We then introduce BinProp, a training-free method that enables these same models to discover reusable chess concepts and use them to explain moves in natural language. BinProp improves explainability for every model tested, nearly tripling the average score compared with direct prompting. Beyond chess, the same method discovers useful concepts in Go, draughts, and shogi without additional tuning.
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.23166 [cs.CL] |
| (or arXiv:2610.23166v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.23166 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Ori Press [view email]
[v1]
Fri, 16 Oct 2026 09:05:17 UTC (9,480 KB)
来源：arXiv cs.CL · arxiv.org