Unconditional Image Generation on ImageNet-1K (train val)
8.9FIDADDP
Evaluation results
| Method | Links | ||
|---|---|---|---|
| ADDPModels=-2026.05 | 8.9 | 95.32 | |
| SorcenBackbone=ViT-B2026.05 | 9.61 | 90.96 | |
| LEASEBackbone=ViT-B2026.05 | 9.62 | 91.78 | |
| MAGEBackbone=ViT-B, Status=reproduced2026.05 | 10.88 | 81.59 | |
| MAGEBackbone=ViT-B2026.05 | 11.1 | 81.17 | |
| IC-GANModels=-2026.05 | 15.6 | 59 | |
| MaskGITModels=-2026.05 | 20.7 | 42.08 | |
| BigBiGANModels=-2026.05 | 21.6 | 27.94 | |
| BigGAN+ ClustModels=-2026.05 | 22 | 23.5 | |
| ADMModels=-2026.05 | 26.2 | 39.7 | |
| HiTModels=-2026.05 | 30.8 | 21.64 | |
| DiTModels=-2026.05 | 30.9 | — | |
| MAGE-CBackbone=ViT-B2026.05 | 31.8 | 37.4 | |
| BigGANModels=-2026.05 | 38.6 | 24.7 |