ResearchBenchmarksInverse Reinforcement Learning on HopperFollow68Normalized PerformanceAIRL9.7624.884055.12May 10, 2026Evaluation ResultsMethodMethodLinksNormalized PerformanceAIRLPhase=TrainingPhase=Training2026.0568TRIRLPhase=TrainingPhase=Training2026.0549TRIRLPhase=RetrainingPhase=Retraining2026.0536NEARPhase=TrainingPhase=Training2026.0522NEARPhase=RetrainingPhase=Retraining2026.0522AIRLPhase=RetrainingPhase=Retraining2026.0512