趣味性具身机器人学习
Playful Agentic Robot Learning
AI总结：
提出RATs框架，让机器人通过自主探索学习可复用技能，在LIBERO-PRO和MolmoSpaces上分别提升20.6和17.0个百分点。
摘要：
当前的具身机器人系统可以编写可执行的代码即策略程序、观察反馈并在多次尝试中修正行为，但它们仍然主要是任务驱动的：可复用技能仅在明确指令后获得。我们研究趣味性具身机器人学习，其中具身编码代理在下游任务到来之前，将自主导向的趣味性作为持续技能学习阶段。我们引入RATs，即专为趣味性技能获取设计的机器人代理团队。在趣味性阶段，RATs提出新颖且可学习的探索性任务，规划并执行机器人代码策略，验证中间进展，诊断失败，通过密集的步骤级反馈进行重试，并将成功执行提炼到持久代码技能库中。在测试时，代理从该冻结库中重用相关技能以帮助解决新任务。在LIBERO-PRO和MolmoSpaces上的实验表明，与无趣味性和随机趣味性基线相比，趣味性学习技能在保留的下游任务上分别提升了20.6和17.0个百分点（相对于CaP-Agent0）。此外，学习到的技能可以通过简单地检索到上下文中插入到其他推理时代码即策略代理中，无需微调基础模型，即可在RoboSuite和真实世界迁移中分别提升8.9和8.8个百分点。
英文摘要：
Current agentic robot systems can write executable Code-as-Policy programs, observe feedback, and revise behavior across multiple attempts, but they remain largely task-driven: reusable skills are acquired only after explicit instructions. We study Playful Agentic Robot Learning, where an embodied coding agent uses self-directed play as a continual skill-learning stage before downstream tasks arrive. We introduce RATs, Robotics Agent Teams designed for play-time skill acquisition. During play, RATs proposes novel yet learnable exploratory tasks, plans and executes robot-code policies, verifies intermediate progress, diagnoses failures, retries with dense, step-level feedback, and distills successful executions into a persistent code skill library. At test time, the agent reuses relevant skills from this frozen library to help solve new tasks. Experiments in LIBERO-PRO and MolmoSpaces show that play-learned skills improve held-out downstream tasks over no-play and random-play baselines, with 20.6 and 17.0 percentage-point gains over CaP-Agent0 on LIBERO-PRO and MolmoSpaces, respectively. Moreover, the learned skills can be plugged into other inference-time Code-as-Policy agents by simply retrieving them into the context, improving RoboSuite and real-world transfer by 8.9 and 8.8 points, respectively, without finetuning the underlying model.
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。
