ResearchBenchmarksMulti-Objective Reinforcement Learning on Deep Sea TreasureMulti-Objective Reinforcement Learning on Deep Sea TreasureFollow22,845.4Hypervolume (HV)PCN-907.96085,258.777111,425.51517,592.2529Nov 27, 2024Feb 21, 2025May 18, 2025Aug 12, 2025Nov 6, 2025Jan 31, 2026Apr 27, 2026Evaluation resultsMethodMethodLinksHypervolume (HV)Expected Utility Metric (EUM)UT ScoreED ScorePCN2024.1122,845.453.82——LCN2024.1122,83853.76——GPI-LS2024.1122,622.853.86——MORL-FB2026.049.92—6.43—Optimal2026.049.92—6.89—Envelope2026.049.91—6.5377PD-MORLinterpolator=withinterpolator=with2026.049.33—0.5269PCN2026.048.71—6.1281SFOLS2026.048.46—5.6525MORL/D2026.045.63—3.8815