Towards Interpretable Deep Reinforcement Learning with Human-Friendly Prototypes.

Eoin M. Kenny, Mycal Tucker, Julie Shah

Browse the full ICLR paper archive.