ResearchBenchmarksOffline Multi-Agent Reinforcement Learning on MPE World (Random)Offline Multi-Agent Reinforcement Learning on MPE World (Random)Follow94.3Average Normalized ScoreSPaCQL-0.8623.84548.5573.255Mar 30, 2026Evaluation resultsMethodMethodLinksAverage Normalized ScoreSPaCQL2026.0394.3PLCQL2026.0384.9CFCQL2026.0368DoF2026.0313.1MACQL2026.0311.7OMAR2026.035.9IQL2026.032.9MA-TD3+BC2026.032.8