AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Guiding Token-Sparse Diffusion Models

Felix Krause, Stefan Andreas Baumann, Johannes Schusterbauer, Olga Grebenkova, Ming Gui, Vincent Tao Hu, Björn Ommer

CVPR 2026 · Image generation › Fast diffusion sampling · approach: Training-free diffusion acceleration · 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 CFG with Sparse GuidanceSiT-XL/2 + routing · ImageNet-256 (400 epochs)FID: 2.57 -> 1.58Improved
use SG_FLOPS config instead of CFGSiT-XL/2 + routing · ImageNet-256 (400 epochs)GFLOPs: 228.84 -> 97.67Improved
replace CFG with SG on a masking-trained modelMaskDiT-style SiT-XL/2 · ImageNet-256 (masking, 160 epochs)FID: 5.82 -> 5.73Improved
replace CFG with SG on a routing-trained modelTREAD-style SiT-XL/2 · ImageNet-256 (routing, 160 epochs)FID: 2.95 -> 2.07Improved
replace CFG with SGTREAD-style SiT-XL/2 · ImageNet-256 (routing, 160 epochs)Inception Score: 233.3 -> 223.4Hurt
replace Sparse Guidance with CFG plus dense finetuningsparsely trained SiT-XL/2 · ImageNet-256FIDImproved
combine SG with AutoGuidanceSiT-XL/2 + routing · ImageNet-256 (400 epochs)FID: 2.95 -> 1.58Improved
add SG instead of ICGSiT-XL/2 + routing · ImageNet-256 (400 epochs)FID: 2.81 -> 1.58Improved
raise inference-time sparsity without guidance (omega = 1.0) (ablation)token-sparse diffusion model · ImageNet-256image qualityHurt
increase guidance scale omega together with sparsity (ablation)SG · ImageNet-256 (FID@5k)FIDNo clear effect
use routing instead of masking as train-time sparsity under SG (ablation)SG · ImageNet-256FID: 5.73 -> 2.07Improved
replace CFG with SGTR-DIT-2.5B · HPSv3overall score: 9.21 -> 9.87Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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