AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Revisiting Autoregressive Models for Generative Image Classification

Ilia Sudakov, Artem Babenko, Dmitry Baranchuk

ECCV 2026 · Image generation › Fast diffusion sampling · approach: Distributional control of diffusion · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
marginalize over multiple token orders (K=20) instead of raster order (ablation)RandAR L/16 · ImageNet-Valtop-1 accuracy: 0.701 -> 0.780Improved
single random order (K=1) instead of raster order (ablation)RandAR · ImageNet-Valtop-1 accuracy: 0.712 vs raster-order better at K=1Hurt
increase number of orders K from 1 to 20 (lower bound) (ablation)RandAR XL/16 · ImageNet-Valtop-1 accuracy: 0.712 -> 0.813Improved
use lower bound E[log p] instead of log E[p] (ablation)RandAR XL/16 · ImageNet-Val (K=20)top-1 accuracy: 0.760 -> 0.813Improved
add latent noise augmentation to LlamaGen tokenizer (ablation)RandAR L/16 · ImageNet-Valtop-1 accuracy: 0.769 -> 0.780Improved
add latent noise augmentation to LlamaGen tokenizer (ablation)RandAR XL/16 · ImageNet-Valtop-1 accuracy: 0.812 -> 0.813No clear effect
use LlamaGen tokenizer instead of MaskGIT (ablation)RandAR L/16 · ImageNet-Atop-1 accuracy: 0.101 -> 0.146Improved
score RandAR as any-order AR instead of masked diffusion (ablation)RandAR-B · ImageNet-Val (K=20)top-1 accuracy: 0.662 -> 0.708Improved
score as AR instead of masked diffusion at K=1 (ablation)RandAR-B · ImageNet-Valtop-1 accuracy: 0.348 -> 0.586Improved
scale RandAR from L/16 to XL/16RandAR · ImageNet-Valtop-1 accuracy: 0.780 -> 0.813Improved
replace DiT classifier with RandARgenerative classifier XL/16 · ImageNet-C Gaussiantop-1 accuracy: 0.505 -> 0.652Improved
replace DiT with RandARgenerative classifier · Camelyon17 OOD (B/16)accuracy: 0.586 -> 0.783Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy