ResearchBenchmarksRobotic Manipulation on HandManipulateBlock Full v0Follow88.5Cumulative RegretACDC88.11690.70893.395.892Mar 2, 2026Evaluation ResultsMethodMethodLinksCumulative RegretTime to Threshold (Epochs)ACDCThreshold (θ)=0.20Threshold (θ)=0.202026.0388.575CHERThreshold (θ)=0.20Threshold (θ)=0.202026.0395.1—DTGSHThreshold (θ)=0.20Threshold (θ)=0.202026.0396.5—HERThreshold (θ)=0.20Threshold (θ)=0.202026.0397.2—HEREBPThreshold (θ)=0.20Threshold (θ)=0.202026.0398.1—