ResearchBenchmarksLong-horizon Manipulation on RLBench SimulatedFollow92.2Success Rate (PutRubbish InBin)SADP77.22481.1128588.888May 16, 2026Evaluation ResultsMethodMethodLinksSuccess Rate (PutRubbish InBin)Success Rate (SlideBlock ToTarget)Success Rate (MeatOff Grill)Success Rate (Open Drawer)Success Rate (Close Drawer)Success Rate (PutItem InDrawer)Average Success RateSADP2026.0592.285.678.97593.971.182.8TARAD2026.0588.98573.378.392.873.982DP32026.0577.866.151.15574.446.161.8