AI Paper Atlas

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

Robustness in Text-Attributed Graph Learning: Insights, Trade-offs, and New Defenses

Runlin Lei, Lu Yi, Mingguo He, Pengyu Qiu, Zhewei Wei, Yongchao Liu, Chuntao Hong

ICLR 2026 · Adversarial robustness and security › Attacks on text and graphs · approach: Graph neural network robustness · 1 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
text-structure trade-off: structure-oriented models (LLaGA, GCN) under structural vs textual attackLLaGA/GCN · TAG benchmarksrobust accuracyMixed
text-oriented models (SFT-neighbor, GraphGPT) under textual poisoningSFT-neighbor · CiteSeeraccuracy drop: 25% drop vs 5%-10% for most GNNsHurt
use advanced text encoders with GNNGuardGNNGuard · structural evasion attacksrobust accuracyImproved
add Guardual adaptive filtering to GNNGuardGNNGuard · structural evasionrobust accuracyImproved
add structural noise injection (-noise) to GraphLLMs (ablation)GraphLLMs · structural evasionrobust accuracyImproved
add textual noise injection (-noisetxt) to GraphLLMs (ablation)GraphLLMs · textual evasionrobust accuracyImproved
combine both noises (-noisefull) in GraphLLMs (ablation)GraphLLMs · structural+textual evasionrobust accuracyNo clear effect
add GNNGuard-style edge similarity filtering (-simf) to GraphLLMs (ablation)GraphLLMs · structural vs textual attacksrobust accuracyMixed
modify SFT prompt for adaptive reliance (-simp) (ablation)GraphLLMs · TAG evasion attacksrobust accuracyNo clear effect
add SFT-auto detection and recovery to SFT-neighborSFT-neighbor · structural and textual attacksrobust accuracyImproved
TextFooler attack with mismatched embedding (BoW or RoBERTa vs MiniLM)GNNs · CoraTextFooler attack effectivenessHurt
add validation to attack evaluationGCN · Cora (HeuristicAttack)accuracy: 61.30 -> 70.33Improved

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