Training model: claude-3-5-haiku-20241022
Inference model: gpt-4o-mini
| Dataset | Validation accuracy | Test accuracy |
|---|---|---|
| espionage | 1.0 | 0.85 |
| potions | 0.9 | 0.6 |
| southgermancredit | 0.5769230769230769 | 0.45132743362831856 |
| timetravel_insurance | 0.8 | 0.85 |
| titanic | 0.7450980392156863 | 0.6274509803921569 |
| wisconsin | 0.4444444444444444 | 0.4696969696969697 |