Train Hard, Fight Easy: Robust Meta Reinforcement Learning

A major challenge of reinforcement learning (RL) in real-world applications is the variation between environments, tasks or clients.