Text-Image Alignment on MS-COCO (val)
35CLIP ScoreNone
Evaluation Results
| Method | Links | |
|---|---|---|
| NoneModel=Z-Image, Reg.=–, Cond.=none, alpha=––2026.03 | 35 | |
| Rimsky et al. (2024)Model=Z-Image, Reg.=–, Cond.=min_max, alpha=1.002026.03 | 34 | |
| CAT(Ours)Model=Z-Image, Reg.=–, Cond.=none, alpha=0.752026.03 | 34 | |
| CAT(Ours)Model=Z-Image, Reg.=0.5, Cond.=none, alpha=1.002026.03 | 33 | |
| NoneModel=Infinity, Reg.=–, Cond.=none, alpha=––2026.03 | 33 | |
| Rimsky et al. (2024)Model=Infinity, Reg.=–, Cond.=none, alpha=0.502026.03 | 32 | |
| CAT(Ours)Model=Infinity, Reg.=0.5, Cond.=min_max, alpha=0.502026.03 | 32 | |
| AffineModel=Z-Image, Reg.=–, Cond.=none, alpha=1.002026.03 | 25 | |
| CAT(Ours)Model=Infinity, Reg.=–, Cond.=ood_mahal., alpha=1.002026.03 | 25 | |
| Rodriguez et al. (2025)Model=Z-Image, Reg.=–, Cond.=min_max, alpha=0.252026.03 | 22 | |
| Rodriguez et al. (2025)Model=Infinity, Reg.=–, Cond.=ood_mahal., alpha=0.252026.03 | 16 | |
| AffineModel=Infinity, Reg.=–, Cond.=mahal., alpha=0.502026.03 | 10 |