Reinforcement Learning Closed 10 sec read DeepMind Made A Superhuman AI For 57 Atari Games Frank May 25, 2020 Two Minute Papers takes a look at the paper “Agent57: Outperforming the Atari Human Benchmark” in the following video.