DexHoldem: 使用 Dexterous Embodied 系统进行德州扑克游戏
DexHoldem: An Agentic Robotics Benchmark for Dexterous Manipulation in Texas Hold'em
AI总结：
本文提出DexHoldem，一个基于ShadowHand的现实世界系统级基准，用于评估德克萨斯扑克的灵巧操作。研究通过14个德克萨斯扑克操作原始技能的1470个远程操作示例，测试了代理在感知、执行和决策路由中的能力。
摘要：
评估基于真实灵巧硬件的具身系统需要超越孤立的原始技能：一个代理必须感知一个变化的桌面场景，选择合适的上下文动作，用灵巧的手执行该动作，并确保场景在后续决策中仍然可用。我们介绍了DexHoldem，一个围绕使用ShadowHand进行德克萨斯扑克灵巧操作构建的现实世界系统级基准。DexHoldem提供了1470个远程操作示例，涵盖14个德克萨斯扑克操作原始技能，一个标准化的物理政策基准，以及一个测试代理是否能够恢复所需结构化游戏状态的代理感知基准。在原始执行方面，π_{0.5}获得了最高的任务完成率（61.2%），而π_{0.5}和π_0在保持场景成功率为47.5%时并列。在代理感知方面，Opus 4.7在严格的问题级准确性（34.3%）方面表现最佳，而GPT 5.5在平均领域准确性（66.8%）方面表现最佳，揭示了孤立视觉子能力与完整路由相关状态恢复之间的差距。最后，我们通过三个案例研究实现了完整的具身代理循环，其中等待、恢复调度、人类帮助请求和重复原始执行揭示了在闭环部署过程中感知和策略错误如何累积。DexHoldem因此在共享物理环境中评估了灵巧桌面执行、代理感知和具身决策路由。项目页面：https://dexholdem.github.io/Dexholdem/.
英文摘要：
Evaluating embodied systems with real dexterous hardware requires more than isolated motor-skill tests: an agent must perceive a changing scene (e.g. a tabletop), choose a context-appropriate action, execute it with a dexterous hand, and leave the scene usable for later decisions. We introduce DexHoldem, a comprehensive real-world benchmark evaluating Texas Hold'em related dexterous manipulations with a ShadowHand. DexHoldem provides 1,470 teleoperated demonstrations across 14 Texas Hold'em manipulation primitives, a standardized physical policy benchmark, and an agentic perception benchmark that tests whether agents can recover the structured game state needed for embodied decision making. On primitive execution, $π_{0.5}$ obtains the highest task completion rate ($61.2\%$), while $π_{0.5}$ and $π_0$ tie on scene-preserving success rate ($47.5\%$). On agentic perception, Opus 5.5 narrowly leads on both strict problem-level accuracy ($49.1\%$) and average field-wise accuracy ($80.6\%$); the gap between the two exposes the distance between isolated visual sub-capabilities and complete routing-relevant state recovery. Finally, we instantiate the full embodied-agent loop with one agent--policy pairing over 33 closed-loop hand-level rollouts, in which only $12.1\%$ of hands complete; retries restore the failed primitive in 12 of 34 dispatches and resolve prolonged execution stalls in three of the four completed hands, which would otherwise have required manual termination. Only one hand completes with neither a retry nor a human-help request. DexHoldem therefore evaluates dexterous tabletop execution, agentic perception, and embodied decision routing in a shared physical setting. Project website: https://dexholdem.github.io/Dexholdem/
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。
