ResearchBenchmarksContinuous Control on reacherSingleFollow1.01Average RewardDPG0.979840.987670.99551.00333Sep 9, 2015Evaluation ResultsMethodMethodLinksAverage RewardBest RewardDPGSetup=original with re...Setup=original with replay buffer and batch normalization2015.091.011.083DDPGObservation Space=pixelsObservation Space=pixels2015.0911.083DDPGObservation Space=low-...Observation Space=low-dimensional2015.090.9811.102