AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Nonparametric Data Attribution for Diffusion Models

Yutian Zhao, Chao Du, Xiaosen Zheng, Tianyu Pang, Min Lin

ICML 2026 · Image generation › Fast diffusion sampling · approach: Diffusion models as perception tools

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
use patch-level analytic-score influence (NDA) instead of CLIP cosine similarityNDA · CIFAR-2 LDS (validation)LDS (%): 8.54 -> 24.88Improved
use NDA instead of D-TRAK gradient-based attributionNDA · CelebA LDSLDS gap to D-TRAK: gap 5.94/2.92 (val/gen)Hurt
combine original and low-resolution patch influence (gamma=0.75) (ablation)NDA · CIFAR-2 validation LDSLDS (%)Improved
use larger patch sizes at higher timesteps (ablation)NDA · CIFAR-2 validation LDSLDS (%)Improved
set top-k patches per training image to 100 (ablation)NDA · CIFAR-2/CelebA validation LDSLDS (%)Improved
extend aggregation timesteps beyond tau_max=500 (to >=600) (ablation)NDA · CIFAR-2 validation LDSLDS (%): drop of ~0.5 at tau_max>=600Hurt
use high-noise timesteps (t>=600) for attribution (ablation)NDA · CIFAR-2 validation LDSLDS (%)Hurt
use NDA instead of CLIP/random for removal in counterfactual retrainingNDA · CIFAR-2 counterfactualmedian l2 distance: 6.68 -> 8.43Improved

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