ResearchBenchmarksOffline Reinforcement Learning on OGBench visual-puzzle-3x3-play (5 tasks)Follow23Success RateValue Flows0.126.061217.94Oct 9, 2025Evaluation ResultsMethodMethodLinksSuccess RateValue FlowsPolicy Type=Flow PoliciesPolicy Type=Flow Policies2025.1023IFQLPolicy Type=Flow PoliciesPolicy Type=Flow Policies2025.1021ReBRACPolicy Type=Gaussian P...Policy Type=Gaussian Policies2025.1020FQLPolicy Type=Flow PoliciesPolicy Type=Flow Policies2025.1020IQNPolicy Type=Flow PoliciesPolicy Type=Flow Policies2025.1019IQLPolicy Type=Gaussian P...Policy Type=Gaussian Policies2025.102FBRACPolicy Type=Flow PoliciesPolicy Type=Flow Policies2025.101