ResearchBenchmarksContinuous Control on fixedReacherDoubleFollow0.921Average RewardDDPG0.264760.435130.60550.77587Sep 9, 2015Evaluation ResultsMethodMethodLinksAverage RewardBest RewardDDPGObservation Space=low-...Observation Space=low-dimensional2015.090.9210.996DDPGObservation Space=pixelsObservation Space=pixels2015.090.8720.943DPGSetup=original with re...Setup=original with replay buffer and batch normalization2015.090.290.995