ResearchBenchmarksShadowHandDoorCloseIn on Bi-DexHandsShadowHandDoorCloseIn on Bi-DexHandsFollow295AUC 0:NPPO233.64249.57265.5281.43Jun 7, 2026Evaluation resultsMethodMethodLinksAUC 0:NTtauFinal ScoreGain AUC (%)PPOTuning=HARBORTuning=HARBOR2026.062953,560,0003123.1PPOTuning=DefaultTuning=Default2026.062875,200,000293—TD3Tuning=HARBORTuning=HARBOR2026.0628114,610,00029618.2SACTuning=DefaultTuning=Default2026.062408,690,000257—TD3Tuning=DefaultTuning=Default2026.06237—250—SACTuning=HARBORTuning=HARBOR2026.0623613,410,000267-1.7