The Dormant Neuron Phenomenon in Deep Reinforcement Learning
This addresses a specific bottleneck in deep reinforcement learning for researchers and practitioners, though it is incremental as it builds on existing methods.
The paper identifies the dormant neuron phenomenon in deep reinforcement learning, where inactive neurons reduce network expressivity, and proposes a method called ReDo that recycles these neurons to improve performance.
In this work we identify the dormant neuron phenomenon in deep reinforcement learning, where an agent's network suffers from an increasing number of inactive neurons, thereby affecting network expressivity. We demonstrate the presence of this phenomenon across a variety of algorithms and environments, and highlight its effect on learning. To address this issue, we propose a simple and effective method (ReDo) that Recycles Dormant neurons throughout training. Our experiments demonstrate that ReDo maintains the expressive power of networks by reducing the number of dormant neurons and results in improved performance.