I think this is an excellent resource. I used to work at DeepMind, and Hado & Matteo are two of the best RL researchers there.