Explaining Agent's Decision-making in a Hierarchical Reinforcement Learning Scenario

TL;DR

This paper explores the use of memory-based explainable reinforcement learning in hierarchical environments, enabling agents to explain their actions at both sub-task and global levels for better transparency.

Contribution

It extends memory-based explainable reinforcement learning to hierarchical tasks, allowing agents to provide success probabilities and explanations across multiple levels.

Findings

Memory-based method effectively computes success probabilities in hierarchical tasks.

Agents can explain actions at both sub-task and overall task levels.

The approach enhances transparency in complex reinforcement learning scenarios.

Abstract

Reinforcement learning is a machine learning approach based on behavioral psychology. It is focused on learning agents that can acquire knowledge and learn to carry out new tasks by interacting with the environment. However, a problem occurs when reinforcement learning is used in critical contexts where the users of the system need to have more information and reliability for the actions executed by an agent. In this regard, explainable reinforcement learning seeks to provide to an agent in training with methods in order to explain its behavior in such a way that users with no experience in machine learning could understand the agent's behavior. One of these is the memory-based explainable reinforcement learning method that is used to compute probabilities of success for each state-action pair using an episodic memory. In this work, we propose to make use of the memory-based explainable reinforcement learning method in a hierarchical environment composed of sub-tasks that need to be first addressed to solve a more complex task. The end goal is to verify if it is possible to provide to the agent the ability to explain its actions in the global task as well as in the sub-tasks. The results obtained showed that it is possible to use the memory-based method in hierarchical environments with high-level tasks and compute the probabilities of success to be used as a basis for explaining the agent's behavior.