Explainable Reinforcement Learning via a Causal World Model

TL;DR

This paper introduces a causal world model for reinforcement learning that enhances explainability by capturing long-term action effects through causal chains, while maintaining high accuracy for effective model-based learning.

Contribution

It presents a novel causal world model that explains long-term effects of actions in RL without prior causal knowledge, improving interpretability and accuracy.

Findings

Model accurately captures causal influence of actions.

Enhances explainability without sacrificing performance.

Applicable to model-based reinforcement learning.

Abstract

Generating explanations for reinforcement learning (RL) is challenging as actions may produce long-term effects on the future. In this paper, we develop a novel framework for explainable RL by learning a causal world model without prior knowledge of the causal structure of the environment. The model captures the influence of actions, allowing us to interpret the long-term effects of actions through causal chains, which present how actions influence environmental variables and finally lead to rewards. Different from most explanatory models which suffer from low accuracy, our model remains accurate while improving explainability, making it applicable in model-based learning. As a result, we demonstrate that our causal model can serve as the bridge between explainability and learning.