ResearchBenchmarksPreference Learning on Lunar Lander SimulatedFollow93.3AlignmentInfogain90.1890.9991.892.61Mar 9, 2026Evaluation ResultsMethodMethodLinksAlignmentRegretInfogain2026.0393.31.9CMA-ES-IG2026.0392.12.3CMA-ES2026.0390.33.7