让机器人自己玩着学技能
机器人现在能自己写代码、执行、纠错，但通常得等人下指令才学新技能。这篇让机器人像小孩一样先自己玩：它自己给自己出题（比如“把方块推到左边”），尝试执行，失败就分析原因重试，成功就把方法存进技能库。之后遇到新任务，直接从库里调技能，不用重新训练。实验显示，玩过的机器人任务成功率比没玩过的高20个百分点，而且技能还能借给别的机器人用。它不是你明天能用上的，但指向一个方向：让机器人通过自我探索积累通用能力，而不是每次从零学。
📄 原文摘要(英文)
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.