ResearchBenchmarksReinforcement Learning on MsPacmanReinforcement Learning on MsPacmanFollow508ReturnNSER424.8446.4468489.6May 10, 2026Evaluation resultsMethodMethodLinksReturnAUCSteps to τNconvEfficiency (η)NSERstrategy=Neuro-Symboli...strategy=Neuro-Symbolic Experience Replay2026.055085.33—6.1437UERstrategy=Uniform Exper...strategy=Uniform Experience Replay2026.054792.39—2.78100CERstrategy=Combined Expe...strategy=Combined Experience Replay2026.054734.72—5.5295PERstrategy=Prioritized E...strategy=Prioritized Experience Replay2026.054285.27—5.8290