Explaining Deep Reinforcement Learning Agents in the Atari Domain through a Surrogate Model

Authors

  • Alexander Sieusahai University of Alberta
  • Matthew Guzdial University of Alberta

Keywords:

Explainable AI, Reinforcement Learning, Atari

Abstract

One major barrier to applications of deep Reinforcement Learning (RL) both inside and outside of games is the lack of explainability. In this paper, we describe a lightweight and effective method to derive explanations for deep RL agents, which we evaluate in the Atari domain. Our method relies on a transformation of the pixel-based input of the RL agent to a symbolic, interpretable input representation. We then train a surrogate model, which is itself interpretable, to replicate the behavior of the target, deep RL agent. Our experiments demonstrate that we can learn an effective surrogate that accurately approximates the underlying decision making of a target agent on a suite of Atari games.

Downloads

Published

2021-10-04

How to Cite

Sieusahai, A., & Guzdial, M. (2021). Explaining Deep Reinforcement Learning Agents in the Atari Domain through a Surrogate Model. Proceedings of the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment, 17(1), 82-90. Retrieved from https://ojs.aaai.org/index.php/AIIDE/article/view/18894