BIOENGINEER

Contrastive Learning Tames Out-of-Distribution Actions in Offline Reinforcement Learning

Researchers at Hanyang University have developed TACCO, a contrastive learning method that explicitly identifies and ...