High-stakes specialized classification on ChemProt (val)
60.59Macro F1EvoPool
Evaluation Results
| Method | Links | |
|---|---|---|
| EvoPoolBackbone=gpt-4o-mini2026.06 | 60.59 | |
| LLM annotationBackbone=gpt-4o-mini2026.06 | 32.12 | |
| AlchemistBackbone=gpt-4o-mini2026.06 | 24.47 | |
| DataSculptBackbone=gpt-4o-mini2026.06 | 20.35 |