AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Watch Your Step: Information Injection in Diffusion Models via Shadow Timestep Embedding

An Huang, Junggab Son, Zuobin Xiong

ICML 2026 · Image generation › Fast diffusion sampling · approach: Diffusion models as perception tools · 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
bind MNIST and Fashion-MNIST to shadow timestepsDDPM · CIFAR-10FID: 24.38 -> 21.82Improved
bind shadow datasets with STEDDPM · CIFAR-10ACC: 73.41% -> 75.65%No clear effect
bind shadow datasets with STEDDPM · CIFAR-10exposure rate: 8.02% -> 9.17%Hurt
bind shadow datasets with STEDDPM · MNISTFID: 1.59 -> 1.85No clear effect
use STE-shadow as backdoor channel versus BadDiffusionDDPM · CIFAR-10FID: 22.53 -> 22.07Improved
use STE-explicit timesteps instead of VillanDiffusionDDPM · CIFAR-10FID: 25.66 -> 21.82Improved
use STE watermark versus ROBINDDPM · watermark robustness avgdetection accuracy: 0.98 -> 0.89Hurt
replace DDPM sampler with DDIM in STE (ablation)STE · CIFAR-10FID: 21.82 -> 22.27Hurt
replace DDPM sampler with DPM-Solver in STE (ablation)STE · CIFAR-10FID: 21.82 -> 37.88Hurt
replace DDPM sampler with DPM-Solver in STE (ablation)STE · MNISTFID: 1.85 -> 16.14Hurt
modify scheduler timestep range instead of extending model embedding (ablation)STE reverse design · multi-dataset generationsample qualityHurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 3 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