Optimizing the Neural Architecture of Reinforcement Learning Agents
This work addresses the problem of manually designing neural network architectures for RL agents, offering an automated solution that could benefit RL researchers and practitioners.
The paper investigates the application of neural architecture search (NAS) methods to optimize the neural network architectures of reinforcement learning (RL) agents. Experiments on the Atari benchmark demonstrate that NAS-found architectures outperform manually selected ones.
Reinforcement learning (RL) enjoyed significant progress over the last years. One of the most important steps forward was the wide application of neural networks. However, architectures of these neural networks are typically constructed manually. In this work, we study recently proposed neural architecture search (NAS) methods for optimizing the architecture of RL agents. We carry out experiments on the Atari benchmark and conclude that modern NAS methods find architectures of RL agents outperforming a manually selected one.