ResearchBenchmarksReinforcement Learning on pendulum_discrete StickyReinforcement Learning on pendulum_discrete StickyFollow-30,641,411.5AUC@THFPS-33,904,356.6816-33,057,245.9133-32,210,135.145-31,363,024.3767Feb 6, 2026Evaluation resultsMethodMethodLinksAUC@TFinal@TDelta % (AUC@T)Delta % (Final@T)HFPS2026.02-30,641,411.5-184.969.3-3.5DQN2026.02-33,778,858.79-178.67——