Multi-Agent Reinforcement Learning on Simple Spread N=3
0.0017CollisionsCG-CMARL
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| CG-CMARLlambda=5.00, Pareto-optimal=true2026.06 | 0.0017 | 0.0006 | 3.92 | |
| CG-CMARLlambda=10.00, Pareto-optimal=false2026.06 | 0.0022 | 0.0007 | 1.33 | |
| CG-CMARLlambda=2.00, Pareto-optimal=true2026.06 | 0.003 | 0.001 | 8.25 | |
| CG-CMARLlambda=1.00, Pareto-optimal=true2026.06 | 0.0034 | 0.0011 | 17.5 | |
| CG-CMARLlambda=0.20, Pareto-optimal=true2026.06 | 0.0037 | 0.0012 | 46.17 | |
| CG-CMARLlambda=0.50, Pareto-optimal=false2026.06 | 0.0042 | 0.0014 | 38.08 | |
| CG-CMARLlambda=0.10, Pareto-optimal=true2026.06 | 0.0052 | 0.0017 | 47.33 | |
| QMIXlocal_ratio=0.02026.06 | 0.0052 | 0.0017 | 7.83 | |
| CG-CMARLlambda=0.05, Pareto-optimal=false2026.06 | 0.0054 | 0.0018 | 46.58 | |
| CG-CMARLlambda=0.00, Pareto-optimal=true2026.06 | 0.0055 | 0.0018 | 47.58 | |
| QMIXlocal_ratio=0.32026.06 | 0.006 | 0.002 | 5.6 | |
| MAPPO-Lagcost_limit=1.02026.06 | 0.0067 | 0.0022 | 9.97 | |
| QMIXlocal_ratio=0.62026.06 | 0.0071 | 0.0024 | 5.83 | |
| MAPPOlocal_ratio=0.62026.06 | 0.0084 | 0.0028 | 10.9 | |
| MAPPOlocal_ratio=0.32026.06 | 0.0126 | 0.0042 | 8 | |
| MAPPO-Lagcost_limit=0.12026.06 | 0.0185 | 0.0062 | 10.53 | |
| MAPPOlocal_ratio=0.02026.06 | 0.0232 | 0.0077 | 9.84 | |
| DCGlocal_ratio=0.62026.06 | 0.0322 | 0.0107 | 24.17 | |
| DCGlocal_ratio=0.32026.06 | 0.0439 | 0.0146 | 23.5 | |
| DCGlocal_ratio=0.02026.06 | 0.045 | 0.015 | 21.58 | |
| IQLlocal_ratio=0.62026.06 | 0.0483 | 0.0161 | 15.67 | |
| IQLlocal_ratio=0.02026.06 | 0.0519 | 0.0173 | 20.58 | |
| IQLlocal_ratio=0.32026.06 | 0.0623 | 0.0208 | 17.42 |