How do “you” model people? Reinforcement learning on humans (arjunbazinga.github.io)