ResearchBenchmarksOffline Reinforcement Learning on Minari 21 tasksFollow48.3Minari Adroit ScoreGFP actor40.29242.37144.4546.529Dec 3, 2025Evaluation ResultsMethodMethodLinksMinari Adroit ScoreMinari Hopper ScoreMinari HalfCheetah ScoreMinari Walker2d ScoreAvg Minari Normalized ScoreGFP actorAlgorithm=GFP, Policy...Algorithm=GFP, Policy Component=actor πθ2025.1248.391.7109.1124.574.1GFP VaBCAlgorithm=GFP, Policy...Algorithm=GFP, Policy Component=VaBC πω2025.1246.191.5103.1122.271.6FQLAlgorithm=FQLAlgorithm=FQL2025.1240.679.697.8121.765.9