Addressing reward bias in Adversarial Imitation Learning with neutral reward functions

Open in new window