Agent57: Outperforming the Atari Human Benchmark

DeepMind’s Agent57 is the first RL agent to outperform humans in all 57 Atari benchmark games. It extends previous algorithms like Never Give Up an...
Back to Top