2 datasets found

The dataset used in the paper is the LunarLander environment, which is a classic control problem. The agent learns to land a lunar lander using human feedback.

The dataset used in the paper is a set of environments for reinforcement learning, including CartPole, Pendulum, and LunarLander.