AI Paper Atlas

Research map / Adversarial robustness and security / Attacks on text and graphs

Adversarial Attack and Defense for Denoising Diffusion Sampling

Zhao-Rong Lai, Xiwen Yuan, Jian Weng

ICML 2026 · Adversarial robustness and security › Attacks on text and graphs · approach: Diffusion-based adversarial purification

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 local variation regularization (ADDDS) to ZOD-MCZOD-MC · CIFAR-10 BNN posterior sampling, PGD10classification accuracy (%): 8.860 -> 9.230Improved
add local variation regularization (ADDDS) to ZOD-MCZOD-MC · CIFAR-10 BNN posterior sampling, cleanclassification accuracy (%): 12.030 -> 29.810Improved
add local variation regularization (ADDDS) to ZOD-MCZOD-MC · CIFAR-10 BNN posterior sampling, FGSMclassification accuracy (%): 8.830 -> 14.970Improved
add local variation regularization (ADDDS) to ZOD-MCZOD-MC · CIFAR-100 BNN posterior sampling, PGD10classification accuracy (%): 0.530 -> 5.990Improved
use ADDDS instead of RDMC on CIFAR-10 under PGD10RDMC · CIFAR-10 BNN posterior sampling, PGD10classification accuracy (%): 8.120 -> 9.230No clear effect
apply a Gaussian perturbation (sigma=0.1) at sampling timeZOD-MC · CIFAR-10 BNN posterior samplingclassification accuracy (%): 12.030 -> 11.390Hurt
apply a Gaussian perturbation (sigma=0.1) at sampling timeproximal sampler · CIFAR-10 BNN posterior samplingclassification accuracy (%): 28.800 -> 21.040Hurt
apply a Gaussian perturbation (sigma=0.1) at sampling timeADDDS · CIFAR-10 BNN posterior samplingclassification accuracy (%): 29.810 -> 29.230No clear effect
apply PGD10 attack at sampling timeADDDS · CIFAR-10 BNN posterior samplingclassification accuracy (%): 29.810 -> 9.230Hurt
add LV regularization to the potential minimization (ablation)potential minimization · Critical point V(x*) vs LVV(x*) and local variation: V 3.9966 -> 4.0512; LV 5.9605e -08 -> 0.0893Mixed
replace first-order Monte Carlo samplers by zeroth-order ZOD-MC with LV defenseADDDS · Gaussian mixtures under PGD10MMD and W2Improved

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