Reset & Distill: A Recipe for Overcoming Negative Transfer in Continual Reinforcement Learning

Open in new window