Multi-Agent Reinforcement Learning on Simple Spread N=4
0.64CollisionsCG-CMARL
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| CG-CMARLlambda=10.00, Pareto-optimal=true2026.06 | 0.64 | 0.11 | 4.38 | |
| CG-CMARLlambda=2.00, Pareto-optimal=true2026.06 | 0.7 | 0.12 | 16.5 | |
| CG-CMARLlambda=5.00, Pareto-optimal=false2026.06 | 0.75 | 0.13 | 6.75 | |
| CG-CMARLlambda=1.00, Pareto-optimal=true2026.06 | 1.03 | 0.17 | 26.81 | |
| MAPPO-Lagcost_limit=0.12026.06 | 1.22 | 0.2 | 7.97 | |
| CG-CMARLlambda=0.20, Pareto-optimal=true2026.06 | 1.37 | 0.23 | 38.5 | |
| MAPPO-Lagcost_limit=1.02026.06 | 1.39 | 0.23 | 11.18 | |
| MAPPOlocal_ratio=0.02026.06 | 1.61 | 0.27 | 10.61 | |
| CG-CMARLlambda=0.50, Pareto-optimal=false2026.06 | 1.62 | 0.27 | 32.06 | |
| CG-CMARLlambda=0.10, Pareto-optimal=false2026.06 | 1.71 | 0.29 | 36.56 | |
| MAPPOlocal_ratio=0.32026.06 | 1.71 | 0.29 | 5.19 | |
| MAPPOlocal_ratio=0.62026.06 | 1.72 | 0.29 | 9.88 | |
| QMIXlocal_ratio=0.32026.06 | 1.77 | 0.3 | 3.95 | |
| QMIXlocal_ratio=0.02026.06 | 1.82 | 0.3 | 3.7 | |
| CG-CMARLlambda=0.05, Pareto-optimal=true2026.06 | 1.86 | 0.31 | 38.94 | |
| QMIXlocal_ratio=0.62026.06 | 1.87 | 0.31 | 4.72 | |
| CG-CMARLlambda=0.00, Pareto-optimal=false2026.06 | 2.43 | 0.41 | 37.5 | |
| IQLlocal_ratio=0.62026.06 | 7.03 | 1.17 | 16.56 | |
| DCGlocal_ratio=0.62026.06 | 7.56 | 1.26 | 23.56 | |
| IQLlocal_ratio=0.02026.06 | 8.37 | 1.4 | 17.19 | |
| DCGlocal_ratio=0.02026.06 | 8.56 | 1.43 | 21.5 | |
| DCGlocal_ratio=0.32026.06 | 8.58 | 1.43 | 24.69 | |
| IQLlocal_ratio=0.32026.06 | 8.68 | 1.45 | 16 |