ResearchBenchmarksOffline Goal-Conditioned Reinforcement Learning on cube-double-play oraclerep v0Follow92Task 1 ScoreIQL-2.6421.9346.571.07Oct 26, 2025Evaluation ResultsMethodMethodLinksTask 1 ScoreTask 2 ScoreTask 3 ScoreTask 4 ScoreTask 5 ScoreOverall ScoreIQL2025.10928485124564IVL2025.1088787584759CRL2025.1077423911735TRL2025.1073233031830FBC2025.101710124QRL2025.10700001BC2025.10600001TDP2025.10600001COE2025.10100000