提出可解释图神经网络，实现快速精准的节点特征贡献分析。
B-cos GNNs: Faithful Explanations through Dynamic Linearity
- 用线性聚合与B-cos变换替代非线性函数，实现动态线性
- 单次前向后向传播即可生成实例级解释，速度远超事后方法
- 在真实和合成数据上解释性能领先，适合需要可解释性的场景
我们提出B-cos GNNs，一种内在可解释的图神经网络，其预测可通过单一输入依赖的线性映射精确分解为每个节点、每项特征的贡献。该模型采用线性（求和式）聚合，并将非线性消息传递和更新函数替换为B-cos变换，从而在任务相关权重与输入间建立有意义的对齐关系，这种对齐直接源于模型的动态线性特性。实例级解释仅需一次前向和反向传播，无需额外解释器、修改学习目标或扰动过程。作为GIN的实例，该方法在预测精度上略有损失，但在多样化的合成与真实世界基准测试中实现了顶尖的可解释性，解释速度比事后基线快数个数量级。
原文摘要 · Abstract (English)
We introduce B-cos GNNs, an inherently explainable class of graph neural networks whose predictions decompose exactly into per-node, per-feature contributions via a single input-dependent linear map. B-cos GNNs use linear (sum-based) aggregation and replace non-linear message and update functions with B-cos transforms. This induces meaningful, task-specific weight-input alignment that is directly accessible through the model's dynamic linearity. Instance-level explanations follow from a single forward and backward pass, requiring no auxiliary explainer, modified learning objective, or perturbation procedure. Instantiated as a GIN, our approach trades small losses in predictive accuracy for state-of-the-art explainability across diverse synthetic and real-world benchmarks, producing explanations orders of magnitude faster than post-hoc baselines.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。