$R^3$: Training Robots to Reason in Natural Language via Reinforcement Learning
Research
Source: arXiv cs.ROPublish time unverified
arXiv:2608.26053v1 Announce Type: new Abstract: Reasoning in language allows foundation models to spend more test-time compute on hard problems, such as those requiring decomposition, constraint tracking, and prediction of future consequences. Whether this mechanism can improve robotic manipulation remains unclear, where long-horizon tasks require tracking partial progress, reasoning about object relations, recovering from mistakes, and steering noisy low-level policies.