UCGM-XL/2 (E2E-VAE + 40 sampling steps, without guidance)

Reported on 2 benchmarks across 1 task · 1 paper

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Medical2 results

Image GenerationonImageNet 256x256
FID· 2025-05-12
1.21
best: 1.06 (SiT-XL/2 + UCGM-S (E2E-VAE + 40 sampling steps + CFG))
Unified Continuous Generative Models arXiv:2505.07447
Image GenerationonImageNet 256x256
NFE· 2025-05-12
40
best: 100 (LightningDiT + UCGM-S (VA-VAE + 50 sampling steps + CFG))
Unified Continuous Generative Models arXiv:2505.07447