Environments for Lifelong Reinforcement Learning
This work tackles the problem of enabling general artificial intelligence through lifelong learning, but it is incremental as it focuses on environment design rather than novel algorithms.
The paper addresses the need for environments that support lifelong reinforcement learning, where agents continuously build skills without forgetting, by reviewing existing environments and proposing future recommendations.
To achieve general artificial intelligence, reinforcement learning (RL) agents should learn not only to optimize returns for one specific task but also to constantly build more complex skills and scaffold their knowledge about the world, without forgetting what has already been learned. In this paper, we discuss the desired characteristics of environments that can support the training and evaluation of lifelong reinforcement learning agents, review existing environments from this perspective, and propose recommendations for devising suitable environments in the future.