Generative Evolution on HotpotQA (val)
63.3Score (%)GEPA
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GEPAModel=Qwen3-8B, Time Budget=30 minutes2026.05 | 63.3 | 1 | |
| CombeeModel=Qwen3-8B, Time Budget=30 minutes, K=102026.05 | 62.5 | 0.94 | |
| FlashEvolveModel=Qwen3-8B, Time Budget=30 minutes2026.05 | 61.7 | 0.88 | |
| CombeeModel=Qwen3-8B, Time Budget=30 minutes, K=402026.05 | 58.6 | 0.63 |