ExRL adapts long-chunk robot policies by learning execution length
A lightweight off-policy RL adapter predicts how long to run each action chunk, cutting the adaptation space versus optimizing the full chunk.
Stories per day · 30 days
Relatedaction-chunks1embodied-ai1environment-generation1exrl1hm3d1llm1navigation1procthor1reinforcement-learning1robotics1
2 stories, newest first · one card per story
Stories tagged #policy-adaptation
From Failures to Learnable Worlds: Policy-Conditioned Environment Generation for Navigation Agents
WEDGE generates navigation training environments from a policy's validated failure modes, beating random and hard-example baselines.