Skip to content

Code for the paper : architectural bias in transport-based generative models - precise asymptotics

Notifications You must be signed in to change notification settings

HugoCui/AE_diffusion

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 
 
 

Repository files navigation

AE_diffusion

Code for the paper : architectural bias in transport-based generative models - precise asymptotics (arXiv)

MNIST

  • Simulations.ipynb provides the code necessary to train the generative model, and produce the samples.
  • Theory.ipynb implements the theoretical characterization of Corollary 2.3 on low-dimensional projections of the generated density, for the example of a trimodal Gaussian mixture target (Fig. 1).

Versions: These notebooks employ Python 3.12 , and Pytorch 2.5.

About

Code for the paper : architectural bias in transport-based generative models - precise asymptotics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published