Catastrophic Interference in Reinforcement Learning: A Solution Based on Context Division and Knowledge Distillation

Open in new window