Researchers have developed a new method to improve the explainability of Graph Neural Networks (GNNs). Existing methods often struggle with a distributional shift between the graphs used for training and those used for explanation, leading to unreliable results. The proposed approach trains a robust surrogate model that provides a faithful proxy for the original model's explanations. This is achieved through an iterative re-weighting strategy that refines explanations while maintaining the validity of soft-mask relaxation, demonstrating consistent improvements in explanation quality across various datasets and GNN backbones. AI
IMPACT Improves the interpretability of GNN models, potentially increasing trust and adoption in critical applications.
RANK_REASON The cluster contains a research paper detailing a new methodology for Graph Neural Networks. [lever_c_demoted from research: ic=1 ai=1.0]
- alphaXiv
- arXiv
- CatalyzeX Code Finder for Papers
- CORE Recommender
- DagsHub
- Gotit.pub
- graph neural networks
- Hugging Face
- IArxiv Recommender
- Influence Flower
- ScienceCast
- Zhuomin Chen
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →