研究图神经网络中特征共存现象，揭示模型设计如何影响可解释性。
Superposition in Graph Neural Networks
- 通过几何分析提取图与节点级特征方向
- 宽度增加导致特征重叠出现相变模式
- 浅层模型易陷入低秩嵌入，适合调试优化
解释图神经网络（GNN）困难，因消息传递混合信号且内部通道很少对应人类概念。本文直接在GNN隐空间研究超叠加现象——多个特征共享同一方向。通过控制实验，以明确图概念为基准，提取两类特征：(i) 图级别类条件质心，(ii) 节点级别线性探测方向，并用基础不变的几何诊断方法分析其结构。在GCN/GIN/GAT模型中发现：增加宽度会引发重叠的相变模式；拓扑结构将重叠信息注入节点特征，池化操作部分重构为任务对齐轴；更尖锐的池化提升轴对齐度并减少通道共享；浅层模型可能陷入亚稳态的低秩嵌入。这些结果将表示几何与具体设计选择（宽度、池化、最终层激活）联系起来，提出更具可解释性的GNN设计路径。
原文摘要 · Abstract (English)
Interpreting graph neural networks (GNNs) is difficult because message passing mixes signals and internal channels rarely align with human concepts. We study superposition, the sharing of directions by multiple features, directly in the latent space of GNNs. Using controlled experiments with unambiguous graph concepts, we extract features as (i) class-conditional centroids at the graph level and (ii) linear-probe directions at the node level, and then analyze their geometry with simple basis-invariant diagnostics. Across GCN/GIN/GAT we find: increasing width produces a phase pattern in overlap; topology imprints overlap onto node-level features that pooling partially remixes into task-aligned graph axes; sharper pooling increases axis alignment and reduces channel sharing; and shallow models can settle into metastable low-rank embeddings. These results connect representational geometry with concrete design choices (width, pooling, and final-layer activations) and suggest practical approaches for more interpretable GNNs.
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。