Hateful Meme Classification On Harmeme
Métriques
AUROC
Accuracy
Résultats
Résultats de performance de divers modèles sur ce benchmark
Tableau comparatif
| Nom du modèle | AUROC | Accuracy |
|---|---|---|
| mapping-memes-to-words-for-multimodal-hateful | 92.83 | 81.64 |
| hate-clipper-multimodal-hateful-meme | 91.87 | 83.90 |
| improved-fine-tuning-of-large-multimodal-1 | 93.2 | 88.1 |
| improved-fine-tuning-of-large-multimodal-1 | 92.9 | 87.7 |
| improving-hateful-memes-detection-via | 91.80 | 87.00 |
| prompting-for-multimodal-hateful-meme | 90.96 | 84.47 |
| disentangling-hate-in-online-memes | 86.39 | 81.24 |