On the Importance of Exploration for Generalization in Reinforcement Learning

Open in new window