Combinatorial Optimization on Euclidean Steiner (test)
97.8Average PerformanceClassical Solver
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Classical SolverBudget=B=162026.05 | 97.8 | — | |
| MEMOIR (GPT-5-mini w/ GPT-5 critic)Budget=B=162026.05 | 81.83 | 99.22 | |
| MEMOIR (GPT-5-mini)Budget=B=162026.05 | 72.43 | 100 | |
| GreedyRefineBudget=B=162026.05 | 71.67 | 99.22 | |
| ReEvoBudget=B=162026.05 | 68.86 | 99.22 | |
| AIDEBudget=B=162026.05 | 63.37 | 99.22 | |
| FunSearchBudget=B=162026.05 | 63.18 | 99.22 | |
| MCTS-AHDBudget=B=162026.05 | 59.17 | 99.22 | |
| o3-mini-highBudget=B=162026.05 | 45.41 | 99.22 | |
| MEMOIR (Qwen2.5-Coder-32B)Budget=B=162026.05 | 31.24 | 75.44 | |
| GPT-5-miniBudget=B=162026.05 | 30.38 | 99.22 | |
| QwQ-32BBudget=B=162026.05 | 27.95 | 100 | |
| MEMOIR (Llama-3.3-70B-Instruct)Budget=B=162026.05 | 5.43 | 99.22 | |
| Llama-3.3-70B-InstructBudget=B=162026.05 | 3.12 | 29.69 | |
| DeepSeek-R1-Distill-Llama-70BBudget=B=162026.05 | 1.56 | 99.22 |