On The Presence of Double-Descent in Deep Reinforcement Learning

Viktor Veselý, Aleksandar Todorov, Matthia Sabatelli

Keywords: Double-Descent, Generalization, Overfitting, Deep Reinforcement Learning

Abstract: The double descent (DD) paradox, where over-parameterized models see generalization improve past the interpolation point, remains largely unexplored in the non-stationary domain of Deep Reinforcement Learning (DRL). We present preliminary evidence that DD exists in model-free DRL, investigating it systematically across varying model capacity using the Actor-Critic framework. We rely on an information-theoretic metric, Policy Entropy, to measure policy uncertainty throughout training. Our results show a clear epoch-wise DD curve; the policy's entrance into the second descent region correlates with a sustained, significant reduction in Policy Entropy. This entropic decay suggests that over-parameterization acts as an implicit regularizer, guiding the policy towards robust, flatter minima in the loss landscape. These findings establish DD as a factor in DRL and provide an information-based mechanism for designing agents that are more general, transferable, and robust.

Serve As Reviewer: ~Viktor_Veselý1, ~Aleksandar_Todorov1

Submission Number: 10