ResearchBenchmarksMulti-Agent Reinforcement Learning on Google Research Football academy_3_vs_1_with_keeperMulti-Agent Reinforcement Learning on Google Research Football academy_3_vs_1_with_keeperFollow430.36Params (K)QLLM423.5144469.7222515.93562.1378Apr 17, 2025Evaluation resultsMethodMethodLinksParams (K)Baseline Average ScoreReduction RateQLLM2025.04430.36575.3625.23QMIX2025.04529.3——QPLEX2025.04543.01——RIIT2025.04568.09——Qatten2025.04601.5——