ResearchBenchmarksOffline Control on Heterogeneous Pendulum Low-Data 100,000 transition stepsFollow-1.39Cumulative RewardAOC-883.3724-654.3962-425.42-196.4438Oct 11, 2023Evaluation ResultsMethodMethodLinksCumulative RewardAOC2023.10-1.39MPC2023.10-1.5MFRL2023.10-4.1Data-Avg-Return2023.10-307.81BC2023.10-422.771NN2023.10-557.07KNN2023.10-849.45