| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| SeeTRUE | VQ2 | Winoground63.5 | 14 | 1mo ago | |
| Winoground (test) | Text Score89.5 | 12 | 1mo ago | ||
| MS-COCO 5K (test) | BLEU32.5 | 8 | 4d ago | ||
| ProtoBias Objects (300 human annotated samples) | SC Score1 | 7 | 1mo ago | ||
| ProtoBias Demography | ProtoScore | SC75 | 7 | 1mo ago | |
| ProtoBias Animals (300 human annotated samples) | SC Score0.89 | 7 | 1mo ago | ||
| T2I-CompBench (val) | LOOP | Image Alignment Reward20.912 | 4 | 1mo ago | |
| MJ-Bench | MT-RL-Judge | Macro F160.59 | 3 | 1mo ago | |
| GenAI-Bench Advanced | CLIP L14 | Alignment Score0.276 | 3 | 1mo ago | |
| GenAI-Bench Basic | CLIP L14 | Alignment Score29.6 | 3 | 1mo ago |
