Counterfactual State Explanations for Reinforcement Learning Agents via Generative Deep Learning

TL;DR

This paper introduces a novel generative deep learning approach to produce counterfactual state explanations for deep reinforcement learning agents in visual environments, improving interpretability for non-expert users.

Contribution

It proposes counterfactual state explanations for RL agents, demonstrating their effectiveness in helping non-experts identify flawed agents through user studies.

Findings

Counterfactual explanations enable non-experts to better identify flawed agents.

Generated counterfactual states are sufficiently realistic for human interpretation.

Counterfactual explanations outperform nearest neighbor baselines in user studies.

Abstract

Counterfactual explanations, which deal with "why not?" scenarios, can provide insightful explanations to an AI agent's behavior. In this work, we focus on generating counterfactual explanations for deep reinforcement learning (RL) agents which operate in visual input environments like Atari. We introduce counterfactual state explanations, a novel example-based approach to counterfactual explanations based on generative deep learning. Specifically, a counterfactual state illustrates what minimal change is needed to an Atari game image such that the agent chooses a different action. We also evaluate the effectiveness of counterfactual states on human participants who are not machine learning experts. Our first user study investigates if humans can discern if the counterfactual state explanations are produced by the actual game or produced by a generative deep learning approach. Our second user study investigates if counterfactual state explanations can help non-expert participants identify a flawed agent; we compare against a baseline approach based on a nearest neighbor explanation which uses images from the actual game. Our results indicate that counterfactual state explanations have sufficient fidelity to the actual game images to enable non-experts to more effectively identify a flawed RL agent compared to the nearest neighbor baseline and to having no explanation at all.