ResearchBenchmarksPolicy Ranking on HopperFollow0.51RegretMOReL0.19042.34774.5056.6623Jun 24, 2026Evaluation ResultsMethodMethodLinksRegretSpearman CorrelationMOReL2026.060.510.57FQE2026.060.540.78LNLLsimArchitecture=best-perf...Architecture=best-performing from Figure 42026.060.590.64LDT2Architecture=best-perf...Architecture=best-performing from Figure 42026.060.60.66ROMI2026.060.640.6VaGraM2026.060.650.46HDTwin2026.060.940.4MOPO2026.068.50.27