AsymFlow uses rank-asymmetric velocity prediction to reach 1.57 FID on ImageNet 256x256 and enables finetuning of latent flow models into superior pixel-space text-to-image generators.
Revisiting diffusion model predictions through dimensionality
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 2years
2026 2verdicts
UNVERDICTED 2roles
background 1polarities
background 1representative citing papers
JLT shows clean-latent prediction outperforms velocity prediction in a matched latent diffusion Transformer, reaching FID-50K 2.50 on ImageNet 256x256.
citing papers explorer
-
Asymmetric Flow Models
AsymFlow uses rank-asymmetric velocity prediction to reach 1.57 FID on ImageNet 256x256 and enables finetuning of latent flow models into superior pixel-space text-to-image generators.
-
JLT: Clean-Latent Prediction in Latent Diffusion Transformers
JLT shows clean-latent prediction outperforms velocity prediction in a matched latent diffusion Transformer, reaching FID-50K 2.50 on ImageNet 256x256.