ResearchBenchmarksOff-policy evaluation on ALFWorld (iter3)Follow0.8RhoADWM-0.0320.1840.40.616Jun 4, 2026Evaluation ResultsMethodMethodLinksRhoCI95 BoundP-ValueADWMn=5n=52026.060.80.110.104ISn=5n=52026.060.7——DRn=5n=52026.060.4——FQEn=5n=52026.060——DMn=5n=52026.060——WISn=5n=52026.060——