Skip to content

Instantly share code, notes, and snippets.

@madebyollin

madebyollin/dc_ae_review.md

Last active August 19, 2025 14:07
Show Gist options

  • Select an option

    Learn more about clone URLs
  • Save madebyollin/0cb99cc86b4ee5394fac9a74e5f9aa63 to your computer and use it in GitHub Desktop.

Select an option

Learn more about clone URLs
Save madebyollin/0cb99cc86b4ee5394fac9a74e5f9aa63 to your computer and use it in GitHub Desktop.
Download ZIP
Reviewing the claims of DC-AE
Raw
dc_ae_review.md

Reviewing the Claims of DC-AE

TL;DR - I think the paper is a good contribution and basically holds up, but Figure 2 seems suspicious and the released repo doesn't include the pieces (AE training code and pretrained 4096-element AEs) that would be needed to make DC-AE practically competitive with SD/SDXL VAEs.

DC-AE is an MIT / Tsinghua / NVIDIA paper about improving generative autoencoders (like the SD VAE) under the high-spatial-compression ratio regime.

I am interested in improved autoencoders, so this gist/thread is my attempt to analyze and review some key claims from the DC-AE paper.

(Disclaimer: I work at NVIDIA in an unrelated org :) - this review is written in my personal capacity as an autoencoder buff).

@forever208
Copy link

forever208 commented Aug 19, 2025

you could go to their OpenReview to ask your questions
https://openreview.net/forum?id=wH8XXUOUZU

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment