Generalizing Behavior via Inverse Reinforcement Learning with Closed-Form Reward Centroids

Re-assign community