ResearchBenchmarksOffline Goal-conditioned Reinforcement Learning on puzzle 3x3-play-oraclerep v0Follow99task1IQL-0.8425.085176.92Oct 26, 2025Evaluation ResultsMethodMethodLinkstask1task2task3task4task5overallIQL2025.10999999989598TRL2025.109999100989999CRL2025.101561125FBC2025.101011123COE2025.10721102TDP2025.10510212BC2025.10411011IVL2025.10422112QRL2025.10300001