On the probability of necessity and sufficiency of explaining Graph Neural Networks: A lower bound optimization approach

• Main Authors: CAI, Ruichu, ZHU, Yuxuan, CHEN, Xuexin, FANG, Yuan, WU, Min, QIAO, Jie, HAO, Zhifeng
• 格式: text
• 語言: English
• 出版: Institutional Knowledge at Singapore Management University 2025
• 主題: Causality; Explainability; Explainable AI; Graph Neural Networks; Interpretability; Numerical Analysis and Scientific Computing; OS and Networks
• 在線閱讀: https://ink.library.smu.edu.sg/sis_research/10111; https://ink.library.smu.edu.sg/context/sis_research/article/11111/viewcontent/Prob_Necessity_GNN_sv.pdf

The explainability of Graph Neural Networks (GNNs) is critical to various GNN applications, yet it remains a significant challenge. A convincing explanation should be both necessary and sufficient simultaneously. However, existing GNN explaining approaches focus on only one of the two aspects, neces...