Offline reinforcement learning under value and density-ratio realizability: The power of gaps.

Jinglin Chen, Nan Jiang

Browse the full UAI paper archive.