ResearchBenchmarksMulti-Objective Reinforcement Learning on mo-swimmer v5Multi-Objective Reinforcement Learning on mo-swimmer v5Follow12,100Hypervolume (HV)Oracle10,85211,17611,50011,824Feb 20, 2026Evaluation resultsMethodMethodLinksHypervolume (HV)Epsilon Multi-Objective (EUM)Velocity of Optimality (VO)OracleReward Sparsity=None (...Reward Sparsity=None (Dense)2026.0212,1009.414.22PRISMReward Sparsity=Extrem...Reward Sparsity=Extreme (first objective)2026.0212,1009.444.24BaselineReward Sparsity=Extrem...Reward Sparsity=Extreme (first objective)2026.0210,9004.11.58