GRAFT: Auditing Graph Neural Networks via Global Feature Attribution

• URL: http://arxiv.org/abs/2605.03377v1
• Date: Tue, 05 May 2026 05:36:31 GMT
• Title: GRAFT: Auditing Graph Neural Networks via Global Feature Attribution
• Abstract summary: Graph Neural Networks (GNNs) achieve strong performance on node classification tasks but remain difficult to interpret.<n>We propose GRAFT, a posthoc global explanation framework that identifies class-level feature importance profiles for GNNs.<n>GRAFT provides a practical and interpretable approach for analysing feature-level behaviour in GNNs, bridging quantitative attribution with human-understandable explanations.
• Score: 0.0
• License: http://creativecommons.org/licenses/by-sa/4.0/
• Abstract: Graph Neural Networks (GNNs) achieve strong performance on node classification tasks but remain difficult to interpret, particularly with respect to which input features drive their predictions. Existing global GNN explainers operate at the structural level identifying recurring subgraph motifs, but none explain model behaviour globally at the level of input node attributes. We propose GRAFT, a posthoc global explanation framework that identifies class-level feature importance profiles for GNNs. The method combines diversity-guided exemplar selection, Integrated Gradients-based attribution, and aggregation to construct a global view of feature influence for each class, which can be further expressed as concise natural language rules using a large language model with self-refinement. We evaluate GRAFT across multiple datasets, architectures, and experimental settings, demonstrating its effectiveness in capturing model-relevant features, supporting bias analysis, and enabling feature-efficient transfer learning. In addition, we introduce a structured human evaluation protocol to assess the interpretability of generated rules along dimensions such as accuracy and usefulness. Our results suggest that GRAFT provides a practical and interpretable approach for analysing feature-level behaviour in GNNs, bridging quantitative attribution with human-understandable explanations.

Related papers

• The Underappreciated Power of Vision Models for Graph Structural Understanding [28.781392309798672]
  Graph Neural Networks operate through bottom-up message-passing, fundamentally differing from human visual perception.<n>We investigate the underappreciated potential of vision models for graph understanding, finding they achieve performance comparable to GNNs on established benchmarks.
  arXiv  Detail & Related papers  (2025-10-27T05:11:44Z)
• TreeX: Generating Global Graphical GNN Explanations via Critical Subtree Extraction [38.99239532650183]
  We propose to unbox GNNs by analyzing and extracting critical subtrees incurred by the inner workings of message passing.<n>By aggregating subtrees in an embedding space with an efficient algorithm, we can make intuitive graphical explanations for Message-Passing GNNs on local, class and global levels.
  arXiv  Detail & Related papers  (2025-03-12T04:36:28Z)
• Graph Self-Supervised Learning with Learnable Structural and Positional Encodings [39.20899720477907]
  We introduce emphGenHopNet, a GNN framework that integrates a $k$-hop message-passing scheme.<n>We also propose a structural- and positional-aware GSSL framework that incorporates topological information throughout the learning process.<n>Our work significantly advances GSSL's capability in distinguishing graphs with similar local structures but different global topologies.
  arXiv  Detail & Related papers  (2025-02-22T14:10:06Z)
• Revisiting Graph Neural Networks on Graph-level Tasks: Comprehensive Experiments, Analysis, and Improvements [54.006506479865344]
  We propose a unified evaluation framework for graph-level Graph Neural Networks (GNNs)<n>This framework provides a standardized setting to evaluate GNNs across diverse datasets.<n>We also propose a novel GNN model with enhanced expressivity and generalization capabilities.
  arXiv  Detail & Related papers  (2025-01-01T08:48:53Z)
• Unveiling Global Interactive Patterns across Graphs: Towards Interpretable Graph Neural Networks [31.29616732552006]
  Graph Neural Networks (GNNs) have emerged as a prominent framework for graph mining. This paper proposes a novel intrinsically interpretable scheme for graph classification. Global Interactive Pattern (GIP) learning introduces learnable global interactive patterns to explicitly interpret decisions.
  arXiv  Detail & Related papers  (2024-07-02T06:31:13Z)
• DGNN: Decoupled Graph Neural Networks with Structural Consistency between Attribute and Graph Embedding Representations [62.04558318166396]
  Graph neural networks (GNNs) demonstrate a robust capability for representation learning on graphs with complex structures. A novel GNNs framework, dubbed Decoupled Graph Neural Networks (DGNN), is introduced to obtain a more comprehensive embedding representation of nodes. Experimental results conducted on several graph benchmark datasets verify DGNN's superiority in node classification task.
  arXiv  Detail & Related papers  (2024-01-28T06:43:13Z)
• GNNEvaluator: Evaluating GNN Performance On Unseen Graphs Without Labels [81.93520935479984]
  We study a new problem, GNN model evaluation, that aims to assess the performance of a specific GNN model trained on labeled and observed graphs. We propose a two-stage GNN model evaluation framework, including (1) DiscGraph set construction and (2) GNNEvaluator training and inference. Under the effective training supervision from the DiscGraph set, GNNEvaluator learns to precisely estimate node classification accuracy of the to-be-evaluated GNN model.
  arXiv  Detail & Related papers  (2023-10-23T05:51:59Z)
• DEGREE: Decomposition Based Explanation For Graph Neural Networks [55.38873296761104]
  We propose DEGREE to provide a faithful explanation for GNN predictions. By decomposing the information generation and aggregation mechanism of GNNs, DEGREE allows tracking the contributions of specific components of the input graph to the final prediction. We also design a subgraph level interpretation algorithm to reveal complex interactions between graph nodes that are overlooked by previous methods.
  arXiv  Detail & Related papers  (2023-05-22T10:29:52Z)
• A Variational Edge Partition Model for Supervised Graph Representation Learning [51.30365677476971]
  This paper introduces a graph generative process to model how the observed edges are generated by aggregating the node interactions over a set of overlapping node communities. We partition each edge into the summation of multiple community-specific weighted edges and use them to define community-specific GNNs. A variational inference framework is proposed to jointly learn a GNN based inference network that partitions the edges into different communities, these community-specific GNNs, and a GNN based predictor that combines community-specific GNNs for the end classification task.
  arXiv  Detail & Related papers  (2022-02-07T14:37:50Z)
• Learnable Structural Semantic Readout for Graph Classification [23.78861906423389]
  We propose structural semantic readout (SSRead) to summarize the node representations at the position-level. SSRead aims to identify structurally-meaningful positions by using the semantic alignment between its nodes and structural prototypes. Our experimental results demonstrate that SSRead significantly improves the classification performance and interpretability of GNN classifiers.
  arXiv  Detail & Related papers  (2021-11-22T20:44:27Z)
• Self-supervised Graph-level Representation Learning with Local and Global Structure [71.45196938842608]
  We propose a unified framework called Local-instance and Global-semantic Learning (GraphLoG) for self-supervised whole-graph representation learning. Besides preserving the local similarities, GraphLoG introduces the hierarchical prototypes to capture the global semantic clusters. An efficient online expectation-maximization (EM) algorithm is further developed for learning the model.
  arXiv  Detail & Related papers  (2021-06-08T05:25:38Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.