AI Paper Atlas

Research map / Adversarial robustness and security / Black-box and transfer attacks

SeRI: Gradient-Free Sensitive Region Identification in Decision-Based Black-Box Attacks

Feiyang Wang, Xingquan Zuo, Hai Huang, Gang Chen, Hangwei Qian

ICLR 2026 · Adversarial robustness and security › Black-box and transfer attacks · approach: Robustness benchmarks and defenses

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 SeRI refinementHSJA · ImageNet VGG19 non-targeted, 2000 queriesavg l2 norm: 8.18 -> 6.47Improved
add SeRI refinementHSJA · ImageNet VGG19 targeted, 2000 queriesavg l2 norm: 72.3 -> 48.3Improved
add SeRI refinementCGBA · ImageNet VGG19 targeted, 2000 queriesavg l2 norm: 77.4 -> 53.0Improved
add SeRI refinementCGBA · ImageNet ViT targeted, 2000 queriesavg l2 norm: 36.6 -> 29.0Improved
add SeRI refinementADBA · ImageNet ViT non-targeted, 10000 queriesavg l2 norm: 3.65 -> 2.39Improved
add SeRI refinementRayS · ImageNet VGG19 non-targeted, 10000 queriesavg l2 norm: 2.72 -> 1.56Improved
use SeRI instead of PAR region sensitivity (ablation)RayS · CIFAR-100 non-targeted, 2000 queriesavg l2 norm: 2.54 -> 1.79Improved
use SeRI instead of PAR region sensitivity (ablation)CGBA · CIFAR-100 non-targeted, 2000 queriesavg l2 norm: 1.71 -> 1.48Improved
use SeRI instead of PAR region sensitivity (ablation)CGBA · ImageNet VGG19 non-targeted, 2000 queriesavg l2 norm: 2.81 -> 2.92Mixed
use SeRI instead of PAR region sensitivity (ablation)CGBA · ImageNet VGG19 targeted, 10000 queriesavg l2 norm: 20.4 -> 21.1Hurt
use 4x4 split instead of 2x2 split (ablation)CGBA+SeRI · ImageNet VGG19 non-targeted, 5000 queriesavg l2 norm: 1.589 -> 1.645No clear effect
use 1x2 split instead of 2x2 split (ablation)CGBA+SeRI · ImageNet VGG19 non-targeted, 2000 queriesavg l2 norm: 3.234 -> 3.292No clear effect

Earlier methods this paper compares with

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