ResearchBenchmarksPolicy optimization on Office World Map 2 Exp 5Follow3,767Average Training StepsQR-MAXRM-71,682.32437,600.59946,883.51,456,166.41Dec 16, 2025Evaluation ResultsMethodMethodLinksAverage Training StepsQR-MAXRM2025.123,767QR-MAX2025.1283,076QRM2025.12438,213R-MAX2025.12723,441UCBVI-sB2025.121,150,000UCBVI-B2025.121,400,000UCBVI-H2025.121,890,000