Research map / Graph learning / GNN architectures
Multi-scale Explainer for Graph Neural Networks
Lutong Wu, Shiying Cheng, Zhiqiang Wang, Jianqing Liang, Peng Song, Xizhao Luo, Jiye Liang
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 multi-scale consistency guidance (ablation) | MSExplainer · six benchmarks, edge ratio 0.7 | prediction accuracy | Improved |
| add single-scale adaptive (contrastive) learning (ablation) | MSExplainer · six benchmarks, edge ratio 0.7 | prediction accuracy | Improved |
| use multiple scales (M=9) instead of single scale (M=1) (ablation) | MSExplainer · Mutagenicity | prediction accuracy | Improved |
| increase number of scales beyond M=9 (ablation) | MSExplainer · Mutagenicity | prediction accuracy | No clear effect |
| set loss ratio lambda1/lambda2 within 1 to 10 instead of 0.01 or 100 (ablation) | MSExplainer · prediction accuracy over scales | prediction accuracy | Improved |
| use MSExplainer instead of GSCExplainer | MSExplainer · Mutagenicity (edge ratio 0.9) | prediction accuracy (%): 87.2 -> 100.0 | Improved |
| use MSExplainer instead of OrphicX | MSExplainer · BA-2Motif (edge ratio 0.5) | prediction accuracy (%): 78.0 -> 84.0 | Improved |
| use MSExplainer instead of MixupExplainer | MSExplainer · NCI1 (edge ratio 0.5) | prediction accuracy (%): 72.5 -> 70.8 | Hurt |
| use MSExplainer instead of GSCExplainer | MSExplainer · Mutagenicity | Edge AUC (%): 91.3 -> 93.0 | Improved |
| use parameter sharing across scales | MSExplainer · training cost | computational complexity | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- OrphicX — OrphicX: A Causality-Inspired Latent Variable Model for Interpreting Graph Neural Networks (CVPR 2022) · better on 5 of 5 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Stratified GNN Explanations through Sufficient Expansion (AAAI 2024 · similarity 0.92)
- Towards Robust Fidelity for Evaluating Explainability of Graph Neural Networks (ICLR 2024 · similarity 0.92)
- Higher Order Structures for Graph Explanations (AAAI 2025 · similarity 0.91)
- EiG-Search: Generating Edge-Induced Subgraphs for GNN Explanation in Linear Time (ICML 2024 · similarity 0.91)
- Beyond Topological Self-Explainable GNNs: A Formal Explainability Perspective (ICML 2025 · similarity 0.90)
- Generating In-Distribution Proxy Graphs for Explaining Graph Neural Networks (ICML 2024 · similarity 0.90)
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.
