ResearchBenchmarksContinuous Control on InvertedDoublePendulum v1 (train)Follow9,355.52Max Average ReturnDDPG-160.13682,310.27414,780.6857,251.0959Feb 26, 2018Evaluation ResultsMethodMethodLinksMax Average ReturnDDPG2018.029,355.52TD32018.029,337.47ACKTR2018.029,081.92PPO2018.028,977.94SAC2018.028,487.15Our DDPG2018.028,369.95TRPO2018.02205.85