AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Visual Implicit Autoregressive Modeling

Pengfei Jiang, Jixiang Luo, Luxi Lin, Zhaohong Huang, Xuelong Li

ICML 2026 · Image generation › Fast diffusion sampling · approach: Distillation for few-step generation · 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
replace deep explicit stack with implicit equilibrium layer (VIAR vs VAR-d30, cfg=2.0)VAR-d30 · ImageNet 256x256sFID: 8.86 -> 7.92Improved
replace deep explicit stack with implicit equilibrium layer (cfg=2.0)VAR-d30 · ImageNet 256x256FID: 2.05 -> 2.35Hurt
replace deep explicit stack with implicit layerVAR-d30 · ImageNet 256x256parameters: 2010.0M -> 770.9MImproved
use most aggressive iteration schedule s4VIAR · ImageNet 256x256 (RTX 4090)throughput: 15.16 -> 32.08 images/sImproved
use most aggressive iteration schedule s4 instead of s1VIAR · ImageNet 256x256 (RTX 4090)FID: 2.16 -> 2.43Hurt
replace VAR-d30 by VIARVAR-d30 · training memoryparameter/gradient memory: 7.49 GB -> 2.87 GBImproved
replace VAR-d30 by VIARVAR-d30 · training memoryoptimizer-state memory: 14.98 GB -> 5.74 GBImproved
use constant 20 iterations instead of constant 10 per scale (ablation)VIAR · ImageNet 256x256FID: 2.16 -> 2.16No clear effect
use constant 5 iterations instead of 10 per scale (ablation)VIAR · ImageNet 256x256FID: 2.16 -> 2.27Hurt
use decreasing schedule (20,5) instead of constant (10,10) (ablation)VIAR · ImageNet 256x256FID: 2.16 -> 2.18Hurt
use decreasing schedule (10,5) instead of constant (10,10) (ablation)VIAR · ImageNet 256x256FID: 2.16 -> 2.22Hurt
use decreasing schedule (20,10) instead of constant (10,10) (ablation)VIAR · ImageNet 256x256FID: 2.16 -> 2.16No clear effect

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