LLMs距离专业扑克玩家还有多远？结合代理工具使用的博弈论推理再探
How Far Are LLMs from Professional Poker Players? Revisiting Game-Theoretic Reasoning with Agentic Tool Use
AI总结：
本文提出ToolPoker框架，通过整合外部求解器和专业解释，提升LLMs在扑克博弈中的推理和游戏表现。
摘要：
随着大型语言模型（LLMs）越来越多地应用于高风险领域，其在不确定性环境下进行战略推理的能力变得至关重要。扑克提供了一个严谨的测试平台，要求不仅有强大的行动能力，还需要有原则性的博弈论推理。在本文中，我们对多个现实的扑克任务进行了系统研究，评估了游戏玩法结果和推理轨迹。我们的分析揭示了LLMs无法与传统算法竞争，并识别出三个反复出现的缺陷：依赖启发式方法、事实误解以及“知道-做”差距，即行动与推理相脱节。初步尝试使用行为克隆和分步强化学习改进了推理风格，但仍然不足以实现准确的博弈论玩法。受这些限制的启发，我们提出了ToolPoker，一种整合工具的推理框架，结合外部求解器以产生GTO一致的动作，以及更精确的专业风格解释。实验表明，ToolPoker在游戏玩法上达到最先进的水平，同时生成的推理轨迹紧密反映了博弈论原则。
英文摘要：
As Large Language Models (LLMs) are increasingly applied in high-stakes domains, their ability to reason strategically under uncertainty becomes critical. Poker provides a rigorous testbed, requiring not only strong actions but also principled, game-theoretic reasoning. In this paper, we conduct a systematic study of LLMs in multiple realistic poker tasks, evaluating both gameplay outcomes and reasoning traces. Our analysis reveals LLMs fail to compete against traditional algorithms and identifies three recurring flaws: reliance on heuristics, factual misunderstandings, and a "knowing-doing" gap where actions diverge from reasoning. An initial attempt with behavior cloning and step-level reinforcement learning improves reasoning style but remains insufficient for accurate game-theoretic play. Motivated by these limitations, we propose ToolPoker, a tool-integrated reasoning framework that combines external solvers for GTO-consistent actions with more precise professional-style explanations. Experiments demonstrate that ToolPoker achieves state-of-the-art gameplay while producing reasoning traces that closely reflect game-theoretic principles.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
