Unifying Bayesian Flow Networks and Diffusion Models through Stochastic Differential Equations

arXiv:2404.15766v1 Announce Type: new
Abstract: Bayesian flow networks (BFNs) iteratively refine the parameters, instead of the samples in diffusion models (DMs), of distributions at various noise levels through Bayesian inference. Owing to its differentiable nature, BFNs are promising in modeling both continuous and discrete data, while simultaneously maintaining fast sampling capabilities. This paper aims to understand and enhance BFNs by connecting them with DMs through stochastic differential equations (SDEs). We identify the linear SDEs corresponding to the noise-addition processes in …

arxiv bayesian cs.ai cs.lg differential diffusion diffusion models flow networks stochastic through type