Skip to content

[NeurIPS 2025] Distilled Decoding 2: One-step Sampling of Image Auto-regressive Models with Conditional Score Distillation

Notifications You must be signed in to change notification settings

imagination-research/Distilled-Decoding-2

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 

Repository files navigation

Distilled-Decoding-2

This is the official repository of the paper "Distilled Decoding 2: One-step Sampling of Image Auto-regressive Models with Conditional Score Distillation" (NeurIPS 2025). Distilled Decoding 2 (DD2) provides a new method of tuning an image Auto-regressive (AR) model into an one-step model.

We are still preparing the code and the arXiv version of the paper, and both will be released in the future.

We also recommend you check out our previous work "Distilled Decoding 1: One-step Sampling of Image Auto-regressive Models with Flow Matching" (ICLR 2025) (code paper Website). If you are interested in our works, we warmly welcome citations and discussions!

About

[NeurIPS 2025] Distilled Decoding 2: One-step Sampling of Image Auto-regressive Models with Conditional Score Distillation

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published