A large-scale empirical study across tokenizers and diffusion backbones identifies Velocity Irreducible Variance (VIV) as one of the most stable predictors of latent diffusion generation quality.
Taming sampling perturbations with variance expansion loss for latent diffusion models
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CV 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
DecQ uses detail-condensing queries on shallow and deep VFM features to improve both reconstruction PSNR and generative convergence/FID in RAEs without fine-tuning the encoder.
citing papers explorer
-
Diffusing in the Right Space: A Systematic Study of Latent Diffusability
A large-scale empirical study across tokenizers and diffusion backbones identifies Velocity Irreducible Variance (VIV) as one of the most stable predictors of latent diffusion generation quality.
-
DecQ: Detail-Condensing Queries for Enhanced Reconstruction and Generation in Representation Autoencoders
DecQ uses detail-condensing queries on shallow and deep VFM features to improve both reconstruction PSNR and generative convergence/FID in RAEs without fine-tuning the encoder.