Everybody Needs a Little HELP: Explaining Graphs via Hierarchical Concepts
Not on the map. Why?
Abstract (arXiv)
Graph neural networks (GNNs) have led to major breakthroughs in a variety of domains such as drug discovery, social network analysis, and travel time estimation. However, they lack interpretability which hinders human trust and thereby deployment to settings with high-stakes decisions. A line of interpretable methods approach this by discovering a small set of relevant concepts as subgraphs in the last GNN layer that together explain the prediction. This can yield oversimplified explanations, failing to explain the interaction between GNN layers. To address this oversight, we provide HELP (Hierarchical Explainable Latent Pooling), a novel, inherently interpretable graph pooling approach that reveals how concepts from different GNN layers compose to new ones in later steps. HELP is more than 1-WL expressive and is the first non-spectral, end-to-end-learnable, hierarchical graph pooling method that can learn to pool a variable number of arbitrary connected components. We empirically demonstrate that it performs on-par with standard GCNs and popular pooling methods in terms of accuracy while yielding explanations that are aligned with expert knowledge in the domains of chemistry and social networks. In addition to a qualitative analysis, we employ concept completeness scores as well as concept conformity, a novel metric to measure the noise in discovered concepts, quantitatively verifying that the discovered concepts are significantly easier to fully understand than those from previous work. Our work represents a first step towards an understanding of graph neural networks that goes beyond a set of concepts from the final layer and instead explains the complex interplay of concepts on different levels.
Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 33 of its 49 references that are in the map. First posted on arXiv 25 Nov 2023.
Builds on
33 of 49 references.
References · 33
- 2016Semi-Supervised Classification with Graph Convolutional Networkscited by 11,361
- 2016Convolutional Neural Networks on Graphs with Fast Localized Spectral Filteringcited by 2,089
- 2018Relational inductive biases, deep learning, and graph networkscited by 1,360
- 2018Hierarchical Graph Representation Learning with Differentiable Poolingcited by 787
- 2019GNNExplainer: Generating Explanations for Graph Neural Networkscited by 737
- 2019Graph U-Netscited by 492
- 2020Parameterized Explainer for Graph Neural Networkcited by 426
- 2019Self-Attention Graph Poolingcited by 421
- 2021On Explainability of Graph Neural Networks via Subgraph Explorationscited by 281
- 2019On Completeness-aware Concept-Based Explanations in Deep Neural Networkscited by 239
- 2019Differentiation of Blackbox Combinatorial Solverscited by 226
- 2020XGNN: Towards Model-Level Explanations of Graph Neural Networkscited by 211
21 more
- 2020PGM-Explainer: Probabilistic Graphical Model Explanations for Graph Neural Networkscited by 211
- 2019Explainability Techniques for Graph Convolutional Networkscited by 161
- 2019ASAP: Adaptive Structure Aware Pooling for Learning Hierarchical Graph Representationscited by 156
- 2020Interpreting Graph Neural Networks for NLP With Differentiable Edge Maskingcited by 153
- 2020Higher-Order Explanations of Graph Neural Networks via Relevant Walkscited by 138
- 2021ETA Prediction with Graph Neural Networks in Google Mapscited by 130
- 2019Learning with Fenchel-Young Lossescited by 119
- 2018Towards Sparse Hierarchical Graph Classifierscited by 100
- 2019Graph Convolutional Networks with EigenPoolingcited by 94
- 2020Learning with Differentiable Perturbed Optimizerscited by 81
- 2021Towards Self-Explainable Graph Neural Networkcited by 75
- 2022Global Explainability of GNNs via Logic Combination of Learned Conceptscited by 65
- 2021Logic Explained Networkscited by 45
- 2017Hybrid Approach of Relation Network and Localized Graph Convolutional Filtering for Breast Cancer Subtype Classificationcited by 40
- 2021GCExplainer: Human-in-the-Loop Concept-based Explanations for Graph Neural Networkscited by 35
- 2022GNNInterpreter: A Probabilistic Generative Model-Level Explanation for Graph Neural Networkscited by 34
- 2022Global Concept-Based Interpretability for Graph Neural Networks via Neuron Analysiscited by 32
- 2022Learning with Combinatorial Optimization Layers: a Probabilistic Approachcited by 29
- 2022Encoding Concepts in Graph Neural Networkscited by 18
- 2019A Non-Negative Factorization approach to node pooling in Graph Convolutional Neural Networkscited by 14
- 2021Algorithmic Concept-based Explainable Reasoningcited by 12
Built on by
Citing papers per year: 2024: 1; 2025: 0; 2026: 1.
2 papers.