Thanks for the information and sharing the paper. I also read it. Agree with you that the "simulation-to-reality gap" would be more critical to the reinforcement learning community.