Continual Reinforcement Learning (CRL) agents struggle to adapt to new situations while retaining prior knowledge, leading to a stability-plasticity trade-of…
机构:McGill
来源:arXiv 2611.07327 | AI4Papers 论文推荐平台