AI Paper Atlas

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

Generic Adversarial Attack Framework Against Graph-based Vertical Federated Learning

Yimin Liu, Peng Jiang, Qi Liu, Liehuang Zhu

AAAI 2026 · Adversarial robustness and security › Attacks on text and graphs · approach: Graph neural network robustness · 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
add generalized embedding learning (GEL) (ablation)SGAC · Ali-CVRattack success rate (%): 40.67 -> 53.50Improved
add domain knowledge transfer (DKT) on top of GEL (ablation)SGAC · Ali-CVRattack success rate (%): 53.50 -> 72.34Improved
add shadow input construction (SIC) on top of GEL and DKT (ablation)SGAC · Ali-CVRattack success rate (%): 72.34 -> 85.55Improved
add node-level saliency masking (ablation)SGAC · Ali-CVRattack success rate (%): 11.67±1.05 drop when removedImproved
add edge-level saliency masking (ablation)SGAC · Ali-CVRattack success rate (%): 12.15±0.67 drop when removedImproved
randomly drop edges instead of saliency-based masking (ablation)SGAC · Ali-CVRattack success rate (%): 10.46±1.52 declineHurt
add same-class shadow additions (ablation)SGAC · Ali-CVRattack success rate (%): 0.64±0.11 gainNo clear effect
increase auxiliary labeled nodes per class from 20 to 120SGAC · Ali-CVRattack success rate (%): 76.54 -> 88.54Improved
increase benign passive parties from 2 to 32SGAC · Ali-CVRattack success rate (%): 82.57 -> 75.27Hurt
use SGAC instead of Hijack surrogate attackHijack · Ali-CVRattack success rate (%): 55.18 -> 85.60Improved
apply gradient filtering with top fraction 0.3 or lowerdefense against SGAC · Ali-CVRASR and clean accuracy: clean accuracy 60.47%Mixed
train server with adversarial training (unseen SGAC variant)defense against SGAC · Ali-CVRattack success rate (%): over 81.17% ASR remainsNo clear effect

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