ResearchBenchmarksPolicy Optimization on Office World Map 4 Exp 6Follow5,630Average Training StepsQR-MAXRM-390,558.42,283,713.34,957,9857,632,256.7Dec 16, 2025Evaluation ResultsMethodMethodLinksAverage Training StepsQR-MAXRM2025.125,630QR-MAX2025.1220,150UCBVI-sB2025.1280,160R-MAX2025.1282,000UCBVI-B2025.1290,030UCBVI-H2025.12101,120QRM2025.129,910,340