Rethinking Adversarial Inverse Reinforcement Learning: From the Angles of Policy Imitation and Transferable Reward Recovery

arXiv:2403.14593v1 Announce Type: new
Abstract: Adversarial inverse reinforcement learning (AIRL) stands as a cornerstone approach in imitation learning. This paper rethinks the two different angles of AIRL: policy imitation and transferable reward recovery. We begin with substituting the built-in algorithm in AIRL with soft actor-critic (SAC) during the policy optimization process to enhance sample efficiency, thanks to the off-policy formulation of SAC and identifiable Markov decision process (MDP) models with respect to AIRL. It indeed exhibits a significant improvement in …

abstract actor actor-critic adversarial algorithm arxiv cs.lg imitation learning paper policy recovery reinforcement reinforcement learning stat.ml type