ResearchBenchmarksControl Task on Inverted Pendulum (test)Follow996.35Average RewardPPO125.3604351.4827577.605803.7273Feb 5, 2026Evaluation ResultsMethodMethodLinksAverage RewardPPOVRER=trueVRER=true2026.02996.35PPOVRER=falseVRER=false2026.02977.48TRPOVRER=trueVRER=true2026.02962.76TRPOVRER=falseVRER=false2026.02916.39A2CVRER=trueVRER=true2026.02737.22A2CVRER=falseVRER=false2026.02158.86