AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Randomized Autoregressive Visual Generation

Qihang Yu, Ju He, Xueqing Deng, Xiaohui Shen, Liang-Chieh Chen

ICCV 2025 · Image generation › Autoregressive and latent generation · approach: Next-scale and token autoregressive images · 105 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
anneal randomness (start 200, end 300) versus pure raster order (ablation)RAR-L · ImageNet-256 (RAR-L)FID: 3.08 -> 2.18Improved
pure random order training versus raster order (ablation)RAR-L · ImageNet-256 (RAR-L)FID: 3.08 -> 3.01No clear effect
anneal randomness (start 200, end 300) versus pure random order (ablation)RAR-L · ImageNet-256 (RAR-L)FID: 3.01 -> 2.18Improved
replace first 100 epochs of raster order with random permutation (start 100, end 100) (ablation)RAR-L · ImageNet-256 (RAR-L)FID: 3.08 -> 2.48Improved
start annealing later (start 300, end 400) instead of 200-300 (ablation)RAR-L · ImageNet-256 (RAR-L)FID: 2.18 -> 2.74Hurt
start annealing at 100 and end at 200 instead of 200-300 (ablation)RAR-L · ImageNet-256 (RAR-L)FID: 2.18 -> 2.28Hurt
use z-curve final scan order instead of row-major (ablation)RAR-L · ImageNet-256 (RAR-L)FID: 2.18 -> 2.29Hurt
use spiral in final scan order instead of row-major (ablation)RAR-L · ImageNet-256 (RAR-L)FID: 2.18 -> 2.50Hurt
use alternate final scan order instead of row-major (ablation)RAR-L · ImageNet-256 (RAR-L)FID: 2.18 -> 2.48Hurt
scale model from RAR-B to RAR-XXLRAR · ImageNet-256FID: 1.95 -> 1.48Improved
use RAR-XL versus MaskBit at similar FIDRAR-XL · ImageNet-256 sampling speedimages/sec: 0.7 -> 8.3Improved
use RAR-XXL versus LlamaGen-3B-384RAR-XXL · ImageNet-256FID: 2.18 -> 1.48Improved

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