Capacitated Vehicle Routing Problem (ACO) on CVRP-ACO N=200 (val)
22.353Mean Objective ValueOptimal
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Optimal2026.05 | 22.353 | 0 | |
| AHD Agent w/SRModel=Qwen3-4B-Instruct-2507, Eval=100, Cost ($)=0.0392026.05 | 28.1 | 25.637 | |
| EOHModel=GPT-4o, Eval=100, Cost ($)=0.4012026.05 | 28.578 | 27.895 | |
| MCTS-AHDModel=DeepSeek-V4-Flash, Eval=100, Cost ($)=0.4642026.05 | 29.2 | 30.656 | |
| EOHModel=DeepSeek-V4-Flash, Eval=100, Cost ($)=0.0262026.05 | 29.203 | 30.659 | |
| CALMModel=Qwen3-4B-Instruct-2507, Eval=150, Cost ($)=0.0752026.05 | 30.243 | 35.298 | |
| AHD AgentModel=Qwen3-4B-Instruct-2507, Eval=12.9, Cost ($)=0.0042026.05 | 30.516 | 36.538 | |
| ReEvoModel=DeepSeek-V4-Flash, Eval=100, Cost ($)=0.0652026.05 | 31.735 | 41.997 | |
| MCTS-AHDModel=GPT-4o, Eval=100, Cost ($)=0.9862026.05 | 32.292 | 44.518 | |
| AHD AgentModel=DeepSeek-V4-Flash, Eval=17.6, Cost ($)=0.0662026.05 | 32.841 | 46.953 | |
| AHD AgentModel=GPT-4o, Eval=7.8, Cost ($)=0.0862026.05 | 44.779 | 100.385 | |
| Baseline heuristic2026.05 | 48.389 | 116.568 | |
| ReEvoModel=GPT-4o, Eval=100, Cost ($)=1.1012026.05 | 52.867 | 209.525 |