ResearchBenchmarksPolicy optimization on Office World Map 3, Exp 5Follow5,806Average Training StepsQR-MAXRM-75,161.76471,370.621,017,9031,564,435.38Dec 16, 2025Evaluation ResultsMethodMethodLinksAverage Training StepsQR-MAXRM2025.125,806QR-MAX2025.12159,597QRM2025.12891,160UCBVI-sB2025.121,410,000R-MAX2025.121,581,301UCBVI-B2025.121,620,000UCBVI-H2025.122,030,000