Algomancer/Bayesian-Flow-Networks

Some question aboud your implement

Closed this issue · 0 comments

The paper mentions that "the network operates on the parameters of a data distribution, rather than on a noisy version
of the data itself", but I don't really understand what it means. I want to know that whether if your code (BFN_MNIST.ipynb) implements this idea? Or does your code implement some other idea of the paper? I didn't really understand how your code is innovative with Unet...

Very much looking forward to your response! And I am very much looking forward to discussing with you further about this article's understanding.