Revisiting Plasticity in Visual Reinforcement Learning: Data, Modules and Training Stages

arXiv:2310.07418v2 Announce Type: replace
Abstract: Plasticity, the ability of a neural network to evolve with new data, is crucial for high-performance and sample-efficient visual reinforcement learning (VRL). Although methods like resetting and regularization can potentially mitigate plasticity loss, the influences of various components within the VRL framework on the agent's plasticity are still poorly understood. In this work, we conduct a systematic empirical exploration focusing on three primary underexplored facets and derive the following insightful conclusions: (1) data augmentation is …

abstract arxiv components cs.ai cs.lg data framework loss modules network neural network performance regularization reinforcement reinforcement learning sample training type visual