Attribute Classification on WhoSaidIt public release
100Accuracy (EN)GPT-4.1
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| GPT-4.1Label=male2026.05 | 100 | 80.8 | 88.9 | 88.9 | 90.9 | 87.8 | |
| GPT-4.1Label=elderly2026.05 | 100 | 91.7 | 89.4 | 93.6 | 90 | 93.2 | |
| GPT-4.1Label=parent2026.05 | 98.8 | 100 | 93.8 | 95.8 | 96.8 | 97.6 | |
| GPT-4.1Label=meat-eater2026.05 | 96 | 93.8 | 93.6 | 95.8 | 92.8 | 94.3 | |
| GPT-4.1Label=serious2026.05 | 93.2 | 90.9 | 83.6 | 89.4 | 92.2 | 90.6 | |
| GPT-4.1Label=avg2026.05 | 90.9 | 89.7 | 87.5 | 88.7 | 90.1 | 89.7 | |
| GPT-4.1Label=adult2026.05 | 90.4 | 88.7 | 71.7 | 85.7 | 87.1 | 86.1 | |
| GPT-4.1Label=female2026.05 | 87.5 | 90.9 | 98 | 80 | 92.3 | 91 | |
| GPT-4.1Label=child2026.05 | 82.6 | 80 | 80 | 78.9 | 81.8 | 80.7 | |
| GPT-4.1Label=vegetarian2026.05 | 69.2 | 90.3 | 88.4 | 89.7 | 86.7 | 85.5 |