ResearchBenchmarksReinforcement Learning on acrobot NoisyFollow-80,111,477.55AUC @ THFPS-82,984,359.3436-82,238,515.0318-81,492,670.72-80,746,826.4082Feb 6, 2026Evaluation ResultsMethodMethodLinksAUC @ TFinal Score @ TPercent Change in AUC @ TPercent Change in Final Score @ THFPS2026.02-80,111,477.55-93.943.315DQN2026.02-82,873,863.89-110.55——