SMAC
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Multi-Agent Reinforcement Learning | SMAC (test) | Win Rate (1c3s5z)100 | 63 | |
| Multi-Agent Reinforcement Learning | SMAC v2 (test) | Win Rate (Protoss 5 Units)84 | 35 | |
| Multi-Agent Reinforcement Learning | SMAC | Win Rate (3m)99.7 | 34 | |
| Exposure Intensity evaluation | SMAC II 1c3s5z (test) | Exposure Intensity2.1 | 24 | |
| Exposure Intensity evaluation | SMAC MMM v2 (test) | Exposure Intensity3.03 | 24 | |
| Exposure Intensity evaluation | SMAC II 8m v2 (test) | Exposure Intensity2.2 | 24 | |
| Exposure Intensity evaluation | SMAC II 1c3s6z vs 1c3s5z v2 (test) | Exposure Intensity2.45 | 24 | |
| Adversary Reward | SMAC 1c3s5z II | Adversary Reward18.87 | 24 | |
| Adversary Reward | SMAC MMM II | Adversary Reward19.16 | 24 | |
| Adversary Reward | SMAC 8m II | Adversary Reward17.07 | 24 | |
| Adversary Reward | SMAC 1c3s6z vs 1c3s5z II | Adversary Reward19.41 | 24 | |
| Multi-Agent Reinforcement Learning | SMAC 3m | Win Rate99.9 | 19 | |
| Multi-Agent Reinforcement Learning | SMAC Avg. v1 | Normalized Win Rate1.006 | 18 | |
| Multi-Agent Reinforcement Learning | SMAC 6h* v1 | Normalized Win Rate104.9 | 18 | |
| Multi-Agent Reinforcement Learning | SMAC 3s5z v1 | Normalized Win Rate99.4 | 18 | |
| Multi-Agent Reinforcement Learning | SMAC 2s3z v1 | Normalized Win Rate100.2 | 18 | |
| Multi-Agent Reinforcement Learning | SMAC 3m v1 | Normalized Win Rate100.9 | 18 | |
| Multi-Agent Reinforcement Learning | SMAC maps | 5m_vs_6m Score59 | 18 | |
| Multi-agent Coordination | SMAC 1o10b_vs_1r | Win Rate65 | 16 | |
| Multi-agent reinforcement learning | SMAC 5v5 full 3x5 grid v2 | Win Rate (Protoss)81 | 15 | |
| Multi-Agent Reinforcement Learning | SMAC terran_10v10 v2 (test) | AUC36.4 | 15 | |
| Multi-Agent Reinforcement Learning | SMAC corridor v2 | Win Rate79.2 | 15 | |
| Multi-Agent Reinforcement Learning | SMAC 5m_vs_6m v2 | Win Rate53.9 | 15 | |
| Multi-Agent Reinforcement Learning | SMAC 3s5z vs 3s6z v2 | Win Rate0.801 | 15 | |
| Multi-Agent Reinforcement Learning | SMAC 3m to 8m | Final Win Rate100 | 13 |
Showing 25 of 295 rows
...