Exploiting Language Instructions for Interpretable and Compositional Reinforcement Learning

Open in new window