Research map / Adversarial robustness and security / Attacks on text and graphs
Not All Low-Pass Filters are Robust in Graph Convolutional Networks
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| add GCN-LFR co-training | VanillaGCN · Cora Nettack-One | accuracy: 0.350 -> 0.800 | Helped |
| add GCN-LFR co-training | VanillaGCN · Cora Mettack | accuracy: 0.701 -> 0.777 | Helped |
| add GCN-LFR co-training | VanillaGCN · Citeseer benign | accuracy: 0.721 -> 0.701 | Hurt |
| add GCN-LFR co-training | JK-Net · Coauthor CS benign | accuracy: 0.927 -> 0.891 | Hurt |
| add GCN-LFR co-training | VanillaGCN · Cora benign | accuracy: 0.815 -> 0.818 | No clear effect |
| use learnable filter F instead of fixed truncated eigenvalues (ablation) | GCN-LFR · Cora benign | accuracy: 0.779 -> 0.824 | Helped |
| use learnable filter F instead of fixed truncated eigenvalues (ablation) | GCN-LFR · Cora Mettack | accuracy: 0.726 -> 0.791 | Helped |
| use alternate instead of joint optimization (ablation) | GCN-LFR · Cora benign test | accuracy: 0.817 -> 0.833 | Helped |
| use alternate instead of joint optimization (ablation) | GCN-LFR · training time | acceleration ratio: x1.74 average | Helped |
| train with only low-frequency net LFR (no GCN branch) (ablation) | GCN-LFR · Cora benign | training accuracy | Hurt |
Closest papers by meaning
- Bounding the Expected Robustness of Graph Neural Networks Subject to Node Feature Attacks (ICLR 2024 · similarity 0.92)
- Towards Inductive Robustness: Distilling and Fostering Wave-Induced Resonance in Transductive GCNs against Graph Adversarial Attacks (AAAI 2024 · similarity 0.90)
- On the Stability of Graph Convolutional Neural Networks: A Probabilistic Perspective (NeurIPS 2025 · similarity 0.90)
- Refining Latent Homophilic Structures over Heterophilic Graphs for Robust Graph Convolution Networks (AAAI 2024 · similarity 0.89)
- Adversarial Robust Generalization of Graph Neural Networks (ICML 2025 · similarity 0.88)
- Robust Graph Condensation via Classification Complexity Mitigation (NeurIPS 2025 · similarity 0.88)
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.