Uncertainty-Aware Unsupervised and Robust Reinforcement Learning

Similar with my defense + future directions + some technical details. See slides here.