Multi-Agent Reinforcement Learning on Simple Spread N=10
1.57CollisionsCG-CMARL
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| CG-CMARLlambda=10.00, Pareto-optimal=true2026.06 | 1.57 | 0.03 | 0.52 | |
| CG-CMARLlambda=5.00, Pareto-optimal=true2026.06 | 1.98 | 0.04 | 1.85 | |
| CG-CMARLlambda=2.00, Pareto-optimal=true2026.06 | 3.43 | 0.08 | 12.24 | |
| CG-CMARLlambda=1.00, Pareto-optimal=true2026.06 | 5.34 | 0.12 | 21.8 | |
| CG-CMARLlambda=0.50, Pareto-optimal=true2026.06 | 8.15 | 0.18 | 24.69 | |
| CG-CMARLlambda=0.20, Pareto-optimal=true2026.06 | 10.74 | 0.24 | 27.64 | |
| CG-CMARLlambda=0.10, Pareto-optimal=false2026.06 | 11.67 | 0.26 | 27.18 | |
| CG-CMARLlambda=0.00, Pareto-optimal=false2026.06 | 12.23 | 0.27 | 27.14 | |
| IQLlocal_ratio=0.02026.06 | 20.87 | 0.46 | 16.73 |