Review for NeurIPS paper: Bayesian Robust Optimization for Imitation Learning

This is a nice paper on robust/safe imitation learning. Could be better to link with Bayesian safe exploration for RL. After further discussion, all reviewers agree that it should be accepted.