Provably Efficient Off-Policy Adversarial Imitation Learning with Convergence Guarantees

Open in new window