AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Beyond Generative Priors: Minority Sampling with JEPA-Guided Diffusion

Sol Park, Soobin Um

ICML 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
add deferred guidance (tau 0.8 instead of 1.0) (ablation)JEPA guidance · MS-COCO prompts, SDv1.5ImageReward: 0.11 -> 0.23Improved
add deferred guidance (tau 0.8 instead of 1.0) (ablation)JEPA guidance · MS-COCO prompts, SDv1.5CLIPScore: 31.26 -> 31.40Improved
increase guidance step size eta (ablation)JEPA guidance · MS-COCO prompts, SDv1.5JEPA-SCORE: -289.79 -> -344.76Improved
increase guidance step size eta (ablation)JEPA guidance · MS-COCO prompts, SDv1.5CLIPScore: 31.44 -> 31.36Hurt
increase RSVD rank k from 9 to 15 (ablation)JEPA guidance · MS-COCO prompts, SDXL-LightningJEPA-SCORE: -344.85 -> -335.28No clear effect
use JEPA guidance with different encoders (ViT-S, ViT-L, MetaCLIP) (ablation)JEPA guidance · MS-COCO prompts, SDv1.5JEPA-SCORE: -289.79 -> -361.52 (ViT-S, eta 0.08)Improved
replace generator-centric minority samplers with JEPA guidanceJEPA guidance · ImageNet 256x256cFID: 32.01 (BnS) -> 18.33Improved
replace generator-centric minority samplers with JEPA guidanceJEPA guidance · CelebA 64x64cFID: 58.27 (MG) -> 8.50Improved
add JEPA-guided minority samples (30K) to classifier trainingResNet-18 classifier · CelebA 40-attribute classificationF1: 0.746 -> 0.775Improved
add JEPA-guided minority samples (30K) to classifier trainingResNet-18 classifier · CelebA 40-attribute classificationaccuracy: 0.898 -> 0.902No clear effect
JEPA guidance vs MinorityPromptJEPA guidance · MS-COCO prompts, SDv1.5ImageReward: 0.24 -> 0.22Hurt

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