| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| FHM (test) | Qwen-3 | Accuracy76.2 | 51 | 2d ago | |
| MAMI (test) | Accuracy84.86 | 51 | 2d ago | ||
| FHM | ALARM | Macro-F175.79 | 49 | 5d ago | |
| MAMI | ALARM | Accuracy85.5 | 33 | 28d ago | |
| Ex-ToxiCN-MM (test) | InternVL3_5-8B | Accuracy96.7 | 30 | 8d ago | |
| HarM | MOMENTA | Accuracy83.82 | 27 | 28d ago | |
| ToxiCN | ALARM | Accuracy79.21 | 16 | 3mo ago | |
| Chinese Harmful Memes | CLIP + MKE | Precision79.76 | 14 | 3mo ago | |
| MemeMind (test) | MemeGuard | Accuracy86.25 | 11 | 2mo ago | |
| GOAT-Bench In-Domain | REPMD | Racism F188.5 | 11 | 3mo ago | |
| Harm-P (test) | EXPLAINHM | Accuracy90.73 | 10 | 3mo ago | |
| Harm-C (test) | EXPLAINHM | Accuracy87 | 10 | 3mo ago | |
| Twitter Temporal-Evolving Memes 2025 (Oct~Dec) | REPMD | F1 Score86.7 | 8 | 3mo ago | |
| Twitter Temporal-Evolving Memes 2025 (Jul~Sep) | REPMD | F1 Score82.4 | 8 | 3mo ago | |
| Twitter Temporal-Evolving Memes 2025 (Apr~Jun) | REPMD | F1 Score82.4 | 8 | 3mo ago | |
| GOAT-Bench (Out-Of-Domain) | REPMD | Racism F187.1 | 7 | 3mo ago |
