ResearchBenchmarksGoal-oriented Robotic Manipulation on FetchPush 2-p v2Follow95.2Normalized ReturnIQL63.68871.86980.0588.231Feb 11, 2026Evaluation ResultsMethodMethodLinksNormalized ReturnIQL2026.0295.2Flex-f-Q2026.0295.2Flex-f-DICE2026.0293Opti-DICE2026.0287.4CQL2026.0281.5TD3BC2026.0264.9