发表机构
TU Berlin; BIFOLD–Berlin Institute for the Foundations of Learning and Data（柏林工业大学; 柏林学习与数据基础研究所）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究针对图神经网络预测缺乏内在解释的问题，提出GDCE-I方法，结合离散扩散反演技术，在统一评估框架下于四个基准测试中大幅优于相关工作，可生成符合领域规则的可解释图反事实解释。
AI 中文摘要
图神经网络（GNN）在化学、生物学和网络分析等领域的图结构数据上实现了出色的预测性能，但它们无法为自身的预测提供内在解释，这限制了其在高风险和安全关键场景中的应用。反事实解释通过揭示能够改变模型预测的最小结构修改来解决这一问题，然而在图数据上生成此类修改十分困难：搜索空间是离散且组合性的，有效答案必须符合节点和边的类别类型，以及分子图场景下的化学价等领域规则。现有解释器要么放弃编辑需符合数据流形的要求，要么放弃覆盖全部编辑空间的搜索。我们提出通过反演的图扩散反事实解释（GDCE-I），兼顾了这两点：带有新型离散反演方案的离散去噪扩散模型，能够利用全部领域编辑空间生成感知分布的编辑。我们还解决了图反事实解释中存在的不完整且不一致的评估问题，推导了一套解释理想特性的框架，并在统一协议下将其应用于所有方法。在四个基准测试中，GDCE-I在该框架定义的指标上大幅优于相关工作；对于分子领域，我们进一步定性表明，GDCE-I能获得可解释的分布内解决方案。
英文摘要
Graph Neural Networks (GNNs) achieve strong predictive performance on graph-structured data across domains such as chemistry, biology, and network analysis, yet they provide no intrinsic explanation of their predictions. This limits their adoption in high-stakes and safety-critical settings. Counterfactual explanations address this by revealing the minimal structural modifications that would change a model's prediction. On graphs, however, such a modification is hard to produce. The search space is discrete and combinatorial, and a valid answer must respect categorical node and edge types together with domain rules such as chemical valency in the case of molecular graphs. Existing explainers give up one of two things. Either edits are not held on the data manifold, or the search does not span the full edit space. We propose Graph Diffusion Counterfactual Explanation via Inversion (GDCE-I), which gives up neither. A discrete denoising diffusion model with a novel discrete inversion scheme enables distribution-aware edits leveraging the whole domain edit space. We further address the incomplete and inconsistent evaluation of graph counterfactuals by deriving a framework of explanation desiderata and applying it to every method under one shared protocol. Across four benchmarks, GDCE-I outperforms related work by a large margin on the defined framework. For the molecular domain, we further qualitatively show that GDCE-I attains interpretable in-distribution solutions.
