arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2410.06950cs.LGcs.AI
收藏

图神经网络的忠实解释

Faithful Interpretation for Graph Neural Networks

King Abdullah University of Science and Technology(阿卜杜拉国王科技大学);Arizona State University(亚利桑那州立大学);The Hong Kong Polytechnic University(香港理工大学);Zhejiang University(浙江大学);Ant Group(蚂蚁集团)

Lijie Hu, Tianhao Huang, Lu Yu, Wanyu Lin, Tianhang Zheng, Di Wang

AI总结:

针对基于注意力的图神经网络可解释性易受扰动影响而不稳定的问题,提出具备四项关键属性的忠实图注意力解释(FGAI)概念及高效实现方法,搭配新评估指标验证其稳定性与解释可靠性。

摘要:

目前,注意力机制在图神经网络(GNNs)中受到越来越多的关注,例如图注意力网络(GATs)和图Transformer(GTs)。这不仅是因为它们带来了值得称道的性能提升,还因为其能够为通常被认为难以理解的模型行为提供更清晰的解释依据。然而,基于注意力的GNN在训练和测试阶段受到各种扰动源(包括额外边或节点等因素)影响时,其可解释性表现出不稳定性。本文中,我们通过提出一种名为忠实图注意力解释(Faithful Graph Attention-based Interpretation, FGAI)的新概念来解决该问题。具体而言,FGAI在解释的稳定性、敏感性以及最终输出分布方面具备四项关键属性。基于这一概念,我们提出了一种获取FGAI的高效方法,该方法可视为对标准基于注意力的GNN的一种即插即用式修改。为验证我们提出的方案,我们引入了两个专为图解释评估定制的新指标。实验结果表明,FGAI在各类扰动和随机性下表现出更优的稳定性,且能保留注意力的可解释性,这使得FGAI成为一种更忠实、更可靠的解释工具。

英文摘要:

Currently, attention mechanisms have garnered increasing attention in Graph Neural Networks (GNNs), such as Graph Attention Networks (GATs) and Graph Transformers (GTs). It is not only due to the commendable boost in performance they offer but also its capacity to provide a more lucid rationale for model behaviors, which are often viewed as inscrutable. However, Attention-based GNNs have demonstrated instability in interpretability when subjected to various sources of perturbations during both training and testing phases, including factors like additional edges or nodes. In this paper, we propose a solution to this problem by introducing a novel notion called Faithful Graph Attention-based Interpretation (FGAI). In particular, FGAI has four crucial properties regarding stability and sensitivity to interpretation and final output distribution. Built upon this notion, we propose an efficient methodology for obtaining FGAI, which can be viewed as an ad hoc modification to the canonical Attention-based GNNs. To validate our proposed solution, we introduce two novel metrics tailored for graph interpretation assessment. Experimental results demonstrate that FGAI exhibits superior stability and preserves the interpretability of attention under various forms of perturbations and randomness, which makes FGAI a more faithful and reliable explanation tool.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑