AI Paper Atlas

Research map / Adversarial robustness and security / Certified and empirical robustness

Natural Jammers: When Non-Robust Features Become Antagonistic

Jiaming Zhang, Fuyao Zhang, Pengjun Xie, Wei Yang Bryan Lim

NeurIPS 2026 · Adversarial robustness and security › Certified and empirical robustness · approach: Robust overfitting and regularization · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Shows that non-robust features can be antagonistic, acting as "Natural Jammers" that mislead standard models on clean, unmodified images, rather than only serving as helpful shortcuts.

Problem

Prior work frames non-robust features as opportunistic shortcuts that help standard models on clean data; the paper asks whether such features can instead actively mislead standard models on clean images.

Approach

Reframes non-robust features as potentially antagonistic: identifies "Natural Jammers," clean unmodified images whose non-robust cues push standard models toward wrong predictions, using disagreement between a standard model and a robust model.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
remove high-frequency components with a low-pass filter (radius 50) (ablation)standard ResNet-50 on the Jammed Set · ImageNet valaccuracy: 0.0% -> 31.8%Improved
remove high-frequency components with a low-pass filter (radius 30) (ablation)standard ResNet-50 on the Spurious Set · ImageNet valaccuracy: 100.0% -> 23.7%Hurt
remove high-frequency components with a low-pass filter (radius 50) (ablation)standard ResNet-50 on the Common Set · ImageNet valaccuracy: 100.0% -> 85.1%Hurt
zoom in with center crop (ratio 1.0 to 0.5) (ablation)standard ResNet-50 on the Jammed Set · ImageNet valaccuracyImproved
zoom in with center crop (ratio 1.0 to 0.5) (ablation)standard ResNet-50 on the Spurious Set · ImageNet valaccuracyHurt
zoom in with center crop (ratio 1.0 to 0.5) (ablation)standard ResNet-50 on the Common Set · ImageNet valaccuracy: 100% -> 79.9%Hurt
graft jammer high frequencies onto a Common-set host image (ablation)standard ResNet-50 · ImageNet valtransferred toxicity: 0.17% transferred vs 99.9% on own hostNo clear effect
apply 2% edge crop and resize (ablation)Jammed Set of standard ResNet-50 · ImageNet valset size |J_T|: 2,140 -> 2,289No clear effect
apply 2% edge crop and resize (ablation)Jammed Set of standard ResNet-50 · ImageNet valIoU with original Jammed Set: 1.00 -> 0.46Hurt
increase crop from 2% to 8% (ablation)Jammed Set of standard ResNet-50 · ImageNet valIoU with original Jammed Set: 0.46 -> 0.35Hurt
use stronger adversarial robustness budget for the robust lens (l2 eps 0.5 to 3.0) (ablation)ResNet-18 quad-set partition · ImageNet valJammed set share: 5.28% -> 3.85%Hurt
change the architecture from ResNet-18 to WRN-50-2 or ViT (ablation)Jammed Set prevalence · ImageNet valJammed set share: 4.90% -> 4.33% (RN-18 vs WRN-50-2, l2 eps=1)No clear effect

Surprising findings

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