Clustered Policy Decision Ranking
This work addresses the interpretability challenge for RL practitioners, but it is incremental as it builds on existing ranking methods.
The paper tackles the problem of interpreting complex reinforcement learning policies by proposing a black-box method that clusters environment states and ranks clusters based on decision importance, comparing it against a prior statistical fault localization approach.
Policies trained via reinforcement learning (RL) are often very complex even for simple tasks. In an episode with n time steps, a policy will make n decisions on actions to take, many of which may appear non-intuitive to the observer. Moreover, it is not clear which of these decisions directly contribute towards achieving the reward and how significant their contribution is. Given a trained policy, we propose a black-box method based on statistical covariance estimation that clusters the states of the environment and ranks each cluster according to the importance of decisions made in its states. We compare our measure against a previous statistical fault localization based ranking procedure.