Deep Reinforcement Learning (John Schulman, OpenAI)

MuJoCo

Emo Todorov (inferred)

“a simulated humanoid robot um in a physics simulator a realistic physics simulator called Moko and uh it has a neural network policy”— John Schulman