Research map / Graph learning / GNN architectures
Learning Conjoint Attentions for Graph Neural Nets
Tiantian He, Yew-Soon Ong, L Bai
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 changed | Where | Measured | Effect |
|---|---|---|---|
| Add matrix-factorization structural intervention to feature attention (CAT-I-MF) | GAT-style graph attention · Cora node classification | Accuracy: 84.44 ± 0.19 -> 85.38 ± 0.16 | Improved |
| Add matrix-factorization structural intervention to feature attention (CAT-I-MF) | GAT-style graph attention · Cite node classification | Accuracy: 72.38 ± 0.50 -> 73.22 ± 0.19 | Improved |
| Add matrix-factorization structural intervention to feature attention (CAT-I-MF) | GAT-style graph attention · Pubmed node classification | Accuracy: 82.42 ± 0.04 -> 83.90 ± 0.24 | Improved |
| Add matrix-factorization structural intervention to feature attention (CAT-I-MF) | GAT-style graph attention · CoauthorCS node classification | Accuracy: 93.60 ± 0.01 -> 93.74 ± 0.14 | Improved |
| Add matrix-factorization structural intervention to feature attention (CAT-I-MF) | GAT-style graph attention · OGB-Arxiv node classification | Accuracy: 72.79 ± 0.09 -> 72.89 ± 0.06 | Improved |
| Add matrix-factorization structural intervention to feature attention (CAT-I-MF) | GAT-style graph attention · Cora node clustering | Accuracy: 81.75 ± 0.26 -> 82.17 ± 0.11 | Improved |
| Add matrix-factorization structural intervention to feature attention (CAT-I-MF) | GAT-style graph attention · Cite node clustering | Accuracy: 70.55 ± 0.85 -> 71.15 ± 0.12 | Improved |
| Add matrix-factorization structural intervention to feature attention (CAT-I-MF) | GAT-style graph attention · Pubmed node clustering | Accuracy: 82.81 ± 0.32 -> 82.77 ± 0.07 | Hurt |
| Add matrix-factorization structural intervention to feature attention (CAT-I-MF) | GAT-style graph attention · CoauthorCS node clustering | Accuracy: 90.16 ± 0.53 -> 90.26 ± 0.22 | Improved |
| Add matrix-factorization structural intervention to feature attention (CAT-I-MF) | GAT-style graph attention · OGB-Arxiv node clustering | Accuracy: 76.91 ± 0.15 -> 77.72 ± 0.07 | Improved |
| Use structural-only attention (MF or SC) instead of feature-only attention (ablation) | GAT attention · All node classification and clustering datasets | Accuracy | Mixed |
| Combine structural and feature correlations with implicit-direction weighting (ablation) | CAT · Node classification and clustering | Accuracy | Improved |
Closest papers by meaning
- Towards Deep Attention in Graph Neural Networks: Problems and Remedies (ICML 2023 · similarity 0.89)
- HONGAT: Graph Attention Networks in the Presence of High-Order Neighbors (AAAI 2024 · similarity 0.86)
- KAA: Kolmogorov-Arnold Attention for Enhancing Attentive Graph Neural Networks (ICLR 2025 · similarity 0.86)
- Graph Attention is Not Always Beneficial: A Theoretical Analysis of Graph Attention Mechanisms via Contextual Stochastic Block Models (ICML 2025 · similarity 0.86)
- Graph External Attention Enhanced Transformer (ICML 2024 · similarity 0.85)
- NAAM: Node-Aware Attention Mechanism for Distilling GNNs-to-MLP (Student Abstract) (AAAI 2025 · similarity 0.85)
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.
