Calibrated Partial Resets: Preventing Policy Collapse in Continual Reinforcement Learning

arXiv:2607.24996v1 Announce Type: new Abstract: Neural networks are hindered by accumulating dormant neurons and loss of expressivity throughout training, particularly in non-stationary data settings, such as continual supervised and reinforcement learning. Recently, neuron resets have been used to maintain gradient flow and restore plasticity. However, full unit reinitialization often sacrifices peak performance and can destabilize training, leading to policy collapse. To preserve plasticity wi...