mlsanityData to 7 Oct 2026

Recognizing Predictive Substructures with Subgraph Information Bottleneck

arXiv 2103.11155 ↗First posted 20 Mar 2021cs.LG cs.AIcited by 28

Not on the map. Why?

Abstract (arXiv)

The emergence of Graph Convolutional Network (GCN) has greatly boosted the progress of graph learning. However, two disturbing factors, noise and redundancy in graph data, and lack of interpretation for prediction results, impede further development of GCN. One solution is to recognize a predictive yet compressed subgraph to get rid of the noise and redundancy and obtain the interpretable part of the graph. This setting of subgraph is similar to the information bottleneck (IB) principle, which is less studied on graph-structured data and GCN. Inspired by the IB principle, we propose a novel subgraph information bottleneck (SIB) framework to recognize such subgraphs, named IB-subgraph. However, the intractability of mutual information and the discrete nature of graph data makes the objective of SIB notoriously hard to optimize. To this end, we introduce a bilevel optimization scheme coupled with a mutual information estimator for irregular graphs. Moreover, we propose a continuous relaxation for subgraph selection with a connectivity loss for stabilization. We further theoretically prove the error bound of our estimation scheme for mutual information and the noise-invariant nature of IB-subgraph. Extensive experiments on graph learning and large-scale point cloud tasks demonstrate the superior property of IB-subgraph.

Key facts. Cited by 28 papers in the map (data through 7 Oct 2026). Builds on 42 of its 68 references that are in the map. First posted on arXiv 20 Mar 2021.

Builds on

42 of 68 references.

References · 42

  1. 2016Semi-Supervised Classification with Graph Convolutional Networkscited by 11,282
  2. 2017Graph Attention Networkscited by 8,854
  3. 2017Inductive Representation Learning on Large Graphscited by 6,284
  4. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,223
  5. 2018How Powerful are Graph Neural Networks?cited by 4,355
  6. 2017Neural Message Passing for Quantum Chemistrycited by 3,460
  7. 2016Categorical Reparameterization with Gumbel-Softmaxcited by 2,708
  8. 2016Deep Variational Information Bottleneckcited by 915
  9. 2018Hierarchical Graph Representation Learning with Differentiable Poolingcited by 786
  10. 2019GNNExplainer: Generating Explanations for Graph Neural Networkscited by 735
  11. 2018MINE: Mutual Information Neural Estimationcited by 680
  12. 2019DropEdge: Towards Deep Graph Convolutional Networks on Node Classificationcited by 679
30 more
  1. 2018FastGCN: Fast Learning with Graph Convolutional Networks via Importance Samplingcited by 583
  2. 2019DeepGCNs: Can GCNs Go as Deep as CNNs?cited by 527
  3. 2019Graph U-Netscited by 489
  4. 2019InfoGraph: Unsupervised and Semi-supervised Graph-Level Representation Learning via Mutual Information Maximizationcited by 441
  5. 2019Self-Attention Graph Poolingcited by 421
  6. 2020Graph Structure Learning for Robust Graph Neural Networkscited by 368
  7. 2020Point-GNN: Graph Neural Network for 3D Object Detection in a Point Cloudcited by 335
  8. 2018Adversarial Attack on Graph Structured Datacited by 300
  9. 2019Adversarial Attacks on Graph Neural Networks via Meta Learningcited by 261
  10. 2019Learning Discrete Structures for Graph Neural Networkscited by 228
  11. 2017Emergence of Invariance and Disentanglement in Deep Representationscited by 213
  12. 2018Adaptive Sampling Towards Fast Graph Representation Learningcited by 180
  13. 2019Understanding Attention and Generalization in Graph Neural Networkscited by 169
  14. 2020Graph Information Bottleneckcited by 168
  15. 2019ASAP: Adaptive Structure Aware Pooling for Learning Hierarchical Graph Representationscited by 156
  16. 2020Graph Information Bottleneck for Subgraph Recognitioncited by 119
  17. 2018Variational Discriminator Bottleneck: Improving Imitation Learning, Inverse RL, and GANs by Constraining Information Flowcited by 103
  18. 2019Grid-GCN for Fast and Scalable Point Cloud Learningcited by 103
  19. 2019Generalization in Reinforcement Learning with Selective Noise Injection and Information Bottleneckcited by 101
  20. 2020Unsupervised Speech Decomposition via Triple Information Bottleneckcited by 87
  21. 2019InfoBot: Transfer and Exploration via the Information Bottleneckcited by 78
  22. 2019Significance-aware Information Bottleneck for Domain Adaptive Semantic Segmentationcited by 74
  23. 2020Subgraph Neural Networkscited by 73
  24. 2019A Restricted Black-box Adversarial Framework Towards Attacking Graph Embedding Modelscited by 68
  25. 2019Learning Efficient Multi-agent Communication: An Information Bottleneck Approachcited by 60
  26. 2019Graph-Revised Convolutional Networkcited by 58
  27. 2019Semi-Supervised Graph Classification: A Hierarchical Graph Perspectivecited by 55
  28. 2019Edge Contraction Pooling for Graph Neural Networkscited by 53
  29. 2021Label-Efficient Point Cloud Semantic Segmentation: An Active Learning Approachcited by 17
  30. 2019Efficient Algorithms for Densest Subgraph Discoverycited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 4; 2022: 4; 2023: 7; 2024: 3; 2025: 5; 2026: 5.

28 papers.

Sort:

By area:

  1. 2021Graph Structure Learning with Variational Information Bottleneckcited by 81
  2. 2021Improving Subgraph Recognition with Variational Graph Information Bottleneckcited by 45
  3. 2023Does Invariant Graph Learning via Environment Augmentation Learn Invariance?cited by 42
  4. 2021Few-shot Backdoor Defense Using Shapley Estimationcited by 40
  5. 2023Thought Propagation: An Analogical Approach to Complex Reasoning with Large Language Modelscited by 37
  6. 2023Mind the Label Shift of Augmentation-based Graph OOD Generalizationcited by 34
  7. 2022A Survey of Trustworthy Graph Learning: Reliability, Explainability, and Privacy Protectioncited by 20
  8. 2025A Comprehensive Survey on Self-Interpretable Neural Networkscited by 18
  9. 2022BrainIB: Interpretable Brain Network-based Psychiatric Diagnosis with Graph Information Bottleneckcited by 17
  10. 2023CI-GNN: A Granger Causality-Inspired Graph Neural Network for Interpretable Brain Network-Based Psychiatric Diagnosiscited by 16
  11. 2022Finding Diverse and Predictable Subgraphs for Graph Domain Generalizationcited by 9
  12. 2024Adaptive Hierarchical Graph Cut for Multi-granularity Out-of-distribution Detectioncited by 8
16 more
  1. 2025MvHo-IB: Multi-View Higher-Order Information Bottleneck for Brain Disorder Diagnosiscited by 8
  2. 2023BatchSampler: Sampling Mini-Batches for Contrastive Learning in Vision, Language, and Graphscited by 7
  3. 2023A Graph is Worth 1-bit Spikes: When Graph Contrastive Learning Meets Spiking Neural Networkscited by 6
  4. 2024Dynamic Graph Information Bottleneckcited by 5
  5. 2021Towards the Explanation of Graph Neural Networks in Digital Pathology with Information Flowscited by 4
  6. 2023Rumor Detection with Diverse Counterfactual Evidencecited by 4
  7. 2026IBCircuit: Towards Holistic Circuit Discovery with Information Bottleneckcited by 3
  8. 2026Modeling Higher-Order Brain Interactions via a Multi-View Information Bottleneck Framework for fMRI-based Psychiatric Diagnosiscited by 3
  9. 2025Information Bottleneck-Guided Heterogeneous Graph Learning for Interpretable Neurodevelopmental Disorder Diagnosiscited by 2
  10. 2025BrainIB++: Leveraging Graph Neural Networks and Information Bottleneck for Functional Brain Biomarkers in Schizophreniacited by 2
  11. 2026PIME: Prototype-based Interpretable MCTS-Enhanced Brain Network Analysis for Disorder Diagnosiscited by 2
  12. 2022Towards Explanation for Unsupervised Graph-Level Representation Learningcited by 1
  13. 2024Discovering robust biomarkers of psychiatric disorders from resting-state functional MRI via graph neural networks: A systematic reviewcited by 1
  14. 2025Structure Matters: Brain Graph Augmentation via Learnable Edge Masking for Data-efficient Psychiatric Diagnosiscited by 0
  15. 2026BlindU: Blind Machine Unlearning without Revealing Erasing Datacited by 0
  16. 2026Subgraph Concept Networks: Concept Levels in Graph Classificationcited by 0