ResearchBenchmarksInverse Reinforcement Learning on Point MazeFollow1.03Normalized PerformanceTRIRL0.250.45250.6550.8575May 10, 2026Evaluation ResultsMethodMethodLinksNormalized PerformanceTRIRLPhase=TrainingPhase=Training2026.051.03TRIRLPhase=RetrainingPhase=Retraining2026.050.98AIRLPhase=TrainingPhase=Training2026.050.45AIRLPhase=RetrainingPhase=Retraining2026.050.35NEARPhase=TrainingPhase=Training2026.050.28NEARPhase=RetrainingPhase=Retraining2026.050.28