Variational Empowerment as Representation Learning for Goal-Based Reinforcement Learning

NeurIPS 2020