arXiv：cs.LG· Abhijeet Krishnan, Colin M. Potts, Arnav Jhala, Harshad Khadilkar, Shirish Karande, Chris Martens·· 4 小时前AI 评分29
让强化学习智能体学会可解释的复杂博弈策略表示
Learning Explainable Representations of Complex Game-playing Strategies
AI 导读
研究者提出一种类似人类认知的方法，训练强化学习智能体把习得的策略综合为基于对局动作序列的可执行程序，从而自动学习下棋和国际象棋类网格环境任务的策略。这些程序化策略能产生有效动作，并可直接从对局数据中学习。
正文
Abstract:As part of learning to play complex games, human players develop develop abstractions for concepts and strategies of gameplay consistent with game rules to improve their performance. These concepts are applied to explain other players' actions, and to inform their own actions in-game. Understanding other players' strategies is a crucial part of such improvement, but requires time and effort. In this paper, we propose a strategy similar to human cognition for training RL agents to synthesize learned strategies and policies as executable procedures based on sequences of gameplay actions. We present methods to automatically learn such programs to play chess and to solve tasks in a grid-based environment. We show that the learned strategies produce effective actions, and can be learned from gameplay data.
| Subjects: | Artificial Intelligence (cs.AI); Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.07638 [cs.AI] |
| (or arXiv:2610.07638v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.07638 arXiv-issued DOI via DataCite (pending registration) |
|
| Journal reference: | Proceedings of the Eleventh Annual Conference on Advances in Cognitive Systems 2024 |
Submission history
From: Abhijeet Krishnan [view email]
[v1]
Tue, 6 Oct 2026 02:28:48 UTC (349 KB)
来源：arXiv：cs.LG · arxiv.org