Reinforcement learning

Decoder[[rl-decoder]]
The Decision and Trajectory Transformer casts the state, action, and reward as a sequence modeling problem.