Preprint
Aug 2026
Diffusion Models for High-Dimensional Clustered Data: Intrinsic-Dimension Adaptivity via Bayesian Classification
This work interprets denoising as a dynamical Bayesian classifier, and proves that the KL error bound depends linearly on the maximum intrinsic dimension of a cluster, up to a logarithmic factor, even when $K$ grows polynomially with $D$.
Yuga Iguchi, P. Fearnhead
· 0 citations