cjlovering / Towards-Interpretable-Reinforcement-Learning-Using-Attention-Augmented-Agents-Replication