Runtime Optimization on AlgoTune
4.716Affine Transform 2DSMCEVOLVE
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| SMCEVOLVEPre-trained model ensemble=gpt-5-mini + gemini-3-flash, Number of seeds=3, Termination strategy=Automatic (ESS-driven)2026.05 | 4.716 | 11.6 | 1.835 | 9.228 | 19.9 | 6.16 | 2.4 | 45.82 | — | |
| SHINKAEVOLVEPre-trained model ensemble=gpt-5-mini + gemini-3-flash, LLM-call budget=500, Number of seeds=3, Termination strategy=Fixed2026.05 | 1.1032 | 1.2682 | 1.26 | 1.1001 | 1.9817 | 1.4119 | 33.8776 | 34.5129 | 500 | |
| REEVOPre-trained model ensemble=gpt-5-mini + gemini-3-flash, LLM-call budget=500, Number of seeds=3, Termination strategy=Fixed2026.05 | 1.0572 | 0.9963 | 1.0688 | 0.9785 | 1.0756 | 1.2854 | 1.4333 | 1.0501 | 500 | |
| OPENEVOLVEPre-trained model ensemble=gpt-5-mini + gemini-3-flash, LLM-call budget=500, Number of seeds=3, Termination strategy=Fixed2026.05 | 0.9957 | 1.0886 | 1.0767 | 1.359 | 1.6226 | 1.812 | 1.6908 | 0.9753 | 500 |