Fetching the paper…

DEAS: DEtached value learning with Action Sequence for Scalable Offline RL · Around