# Explaining Reinforcement Learning to Mere Mortals: An Empirical Study

## TL;DR

This study investigates how different explanation methods, including saliency maps and reward-decomposition bars, affect non-experts' understanding of reinforcement learning agents in a game setting.

## Contribution

It provides empirical evidence on the effectiveness of combined explanation techniques in improving mental models of RL for non-experts.

## Key findings

- Combined explanations significantly improved understanding.
- Saliency maps alone were less effective.
- Qualitative insights suggest areas for further research.

## Abstract

We present a user study to investigate the impact of explanations on non-experts' understanding of reinforcement learning (RL) agents. We investigate both a common RL visualization, saliency maps (the focus of attention), and a more recent explanation type, reward-decomposition bars (predictions of future types of rewards). We designed a 124 participant, four-treatment experiment to compare participants' mental models of an RL agent in a simple Real-Time Strategy (RTS) game. Our results show that the combination of both saliency and reward bars were needed to achieve a statistically significant improvement in mental model score over the control. In addition, our qualitative analysis of the data reveals a number of effects for further study.