DexHoldem：以靈巧具身系統進行德州撲克對局

摘要

在真实灵巧硬件上评估具身系统，所需的不只是孤立的原始技能：智能体必须感知变化中的桌面场景，选择符合情境的动作，以灵巧手执行该动作，并让场景保持可应用于后续决策的状态。我们提出 DexHoldem，这是一个基于 ShadowHand 执行德州扑克灵巧操作的实际系统级基准测试。DexHoldem 提供了涵盖 14 种德州扑克操作原语的 1,470 组遥操作示范、一个标准化的物理策略基准，以及一个测试智能体能否恢复具身决策所需结构化游戏状态的智能体感知基准。在原始执行方面，π_{0.5} 获得最高的任务完成率（61.2%），而 π_{0.5} 与 π_0 在场景保护成功率上持平（47.5%）。在智能体感知方面，Opus 4.7 获得最佳严格问题级准确率（34.3%），GPT 5.5 则在平均字段级准确率上最优（66.8%），这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例化了完整的具身智能体循环，其中等待、恢复调度、请求人工帮助以及重复执行原语，揭示了封闭式部署中感知与策略误差如何累积。因此，DexHoldem 在同一物理设定下评估了灵巧桌面执行、智能体感知以及具身决策路由。项目页面：https://dexholdem.github.io/Dexholdem/。

English

Evaluating embodied systems on real dexterous hardware requires more than isolated primitive skills: an agent must perceive a changing tabletop scene, choose a context-appropriate action, execute it with a dexterous hand, and leave the scene usable for later decisions. We introduce DexHoldem, a real-world system-level benchmark built around Texas Hold'em dexterous manipulation 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 4.7 obtains the best strict problem-level accuracy (34.3%), while GPT 5.5 obtains the best average field-wise accuracy (66.8%), exposing a gap between isolated visual sub-capabilities and complete routing-relevant state recovery. Finally, we instantiate the full embodied-agent loop in three case studies, where waiting, recovery dispatches, human-help requests, and repeated primitive execution reveal how perception and policy errors accumulate during closed-loop deployment. DexHoldem therefore evaluates dexterous tabletop execution, agentic perception, and embodied decision routing in a shared physical setting. Project page: https://dexholdem.github.io/Dexholdem/.