Multi-Agent Reinforcement Learning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
6.4Convergence Steps (×10^5)
7
Jun 25, 2026
80,000Environment Interactions (x10^5)
7
Jun 25, 2026
0.99Steps to 80% Reward
7
Jun 24, 2026
0.89Environment Steps to 80% Reward
7
Jun 24, 2026
0.62Environment Steps to 80% Reward
7
Jun 24, 2026
62.6Win Rate
7
Jun 4, 2026
72.1Win Rate
7
Jun 4, 2026
94.5Win Rate
7
Jun 4, 2026
100Win Rate
7
Jun 15, 2026
97Win Rate: Pass and Shoot
7
Apr 8, 2026
89Protoss 5v5 Win Rate
7
Apr 8, 2026
6Best Outcome
7
Apr 8, 2026
0.974Win Rate
7
Feb 26, 2026
74.3Win Rate
7
Feb 26, 2026
97.8Win Rate
7
Feb 26, 2026
94.8Win Rate
7
Feb 26, 2026
0.872Win Rate
7
Feb 26, 2026
100Win Rate
7
Feb 26, 2026
100Win Rate
7
Apr 13, 2026
-6.8Avg Episode Reward
7
Feb 26, 2026
-5.8Avg Episode Cumulative Reward
7
Feb 26, 2026
62.1Avg Episode Cumulative Reward
7
Feb 26, 2026
76.9Avg Episode Cumulative Reward
7
Feb 26, 2026
-128Cumulative Reward
7
Feb 26, 2026
-129.4Cumulative Reward
7
Feb 26, 2026