On-Robot Reinforcement Learning with Goal-Contrastive Rewards

Open in new window