AI Paper Atlas

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

Does a Hybrid Space-Aware Randomized Defense Improve Empirical and Certified Adversarial Robustness?

Joy Dhar, Manish Kumar Pandey, Behzad Bozorgtabar, Nayyar Zaidi, Wenyu Zhang, Wei-Hong Li, Tingting Mu, Dwarikanath Mahapatra, Mahsa Baktashmotlagh, Trung Le, Chen Chen, Sajib Mistry and 9 more

ICML 2026 · Adversarial robustness and security › Certified and empirical robustness · approach: Provable defenses and their limits

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 RWAN and SANI hybrid stochasticity over ARSHySCAN · ImageNet and CIFAR-10 certified, r>0certified accuracy: up to 9.6 percentage pointsImproved
add hybrid space-aware stochasticity over HyCASHySCAN · ImageNet and CIFAR-10 certified, r>0certified accuracy: up to 1.9 percentage pointsImproved
add attention-noise hybrid defense over ARS and HyCASHySCAN · NIH-CXR / HAM10000 certified (sigma=1.0, r=1.0)certified accuracy: 34.1 -> 42.9 and 41.4 -> 42.9 (NIH-CXR); 34.6 -> 39.5 and 38.5 -> 39.5 (HAM10000)Improved
raise smoothing noise sigma from 0.25 to 0.50 (ablation)HySCAN · ImageNet certified, r=2.0certified accuracy: 7.15 -> 26.2Improved
raise smoothing noise sigma from 0.25 to 0.50 (ablation)HySCAN · ImageNet certified, r=1.0certified accuracy: 42.3 -> 42.3No clear effect
raise smoothing noise sigma from 0.25 to 0.50 (ablation)HySCAN · CIFAR-10 certified, r=0.75 and r=2.0certified accuracy: 45.5 -> 45.8 (r=0.75); 9.94 -> 13.9 (r=2.0)Mixed
use HySCAN instead of CTRW and other empirical defensesHySCAN · four medical datasets, APGD-20 and AA-20robust accuracy: +2.1 to +5.0 percentage pointsImproved
use HySCAN instead of DCSHySCAN · EyePACS AA-20 (eps=8/255)robust accuracy: up to +7.8 pointsImproved
use HySCAN under strong PGD at eps=0.08HySCAN · CIFAR-10 PGDrobust accuracy: about 9 point advantageImproved
use HySCAN with more PGD stepsHySCAN · CIFAR-100 PGD at 100 stepsrobust accuracy: about 2 point marginImproved
use HySCAN against HyCAS on clean accuracyHySCAN · ImageNet certified (sigma=0.25, r=0)clean accuracy: 72.3 -> 71.9Hurt

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