mlsanityData to 14 Oct 2026

GlanceNets: Interpretabile, Leak-proof Concept-based Models

arXiv 2205.15612 ↗First posted 31 May 2022cs.LGcited by 73

Not on the map. Why?

Abstract (arXiv)

There is growing interest in concept-based models (CBMs) that combine high-performance and interpretability by acquiring and reasoning with a vocabulary of high-level concepts. A key requirement is that the concepts be interpretable. Existing CBMs tackle this desideratum using a variety of heuristics based on unclear notions of interpretability, and fail to acquire concepts with the intended semantics. We address this by providing a clear definition of interpretability in terms of alignment between the model's representation and an underlying data generation process, and introduce GlanceNets, a new CBM that exploits techniques from disentangled representation learning and open-set recognition to achieve alignment, thus improving the interpretability of the learned concepts. We show that GlanceNets, paired with concept-level supervision, achieve better alignment than state-of-the-art approaches while preventing spurious information from unintendedly leaking into the learned concepts.

Key facts. Cited by 73 papers in the map (data through 14 Oct 2026). Builds on 49 of its 81 references that are in the map. First posted on arXiv 31 May 2022.

Builds on

49 of 81 references.

References · 49

  1. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  2. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  3. 2014Deep Learning Face Attributes in the Wildcited by 3,693
  4. 2017Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)cited by 1,319
  5. 2017Explanation in Artificial Intelligence: Insights from the Social Sciencescited by 1,206
  6. 2020Concept Bottleneck Modelscited by 1,032
  7. 2019An Introduction to Variational Autoencoderscited by 1,000
  8. 2018A Survey Of Methods For Explaining Black Box Modelscited by 973
  9. 2018Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representationscited by 941
  10. 2016InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Netscited by 860
  11. 2018This Looks Like That: Deep Learning for Interpretable Image Recognitioncited by 737
  12. 2018Towards Robust Interpretability with Self-Explaining Neural Networkscited by 505
37 more
  1. 2019Variational Autoencoders and Nonlinear ICA: A Unifying Frameworkcited by 503
  2. 2017Wasserstein Auto-Encoderscited by 373
  3. 2017Variational Inference of Disentangled Latent Concepts from Unlabeled Observationscited by 241
  4. 2020Weakly-Supervised Disentanglement Without Compromisescited by 239
  5. 2017Deep Learning for Case-Based Reasoning through Prototypes: A Neural Network that Explains Its Predictionscited by 234
  6. 2020Concept Whitening for Interpretable Image Recognitioncited by 225
  7. 2019Explanations can be manipulated and geometry is to blamecited by 216
  8. 2020Neural Prototype Trees for Interpretable Fine-grained Image Recognitioncited by 214
  9. 2020Evaluating Explainable AI: Which Algorithmic Explanations Help Users Predict Model Behavior?cited by 190
  10. 2018Isolating Sources of Disentanglement in Variational Autoencoderscited by 167
  11. 2019From Variational to Deterministic Autoencoderscited by 133
  12. 2021Promises and Pitfalls of Black-Box Concept Learning Modelscited by 131
  13. 2015Deep Convolutional Inverse Graphics Networkcited by 121
  14. 2021Do Concept Bottleneck Models Learn as Intended?cited by 116
  15. 2018Robustly Disentangled Causal Mechanisms: Validating Deep Representations for Interventional Robustnesscited by 108
  16. 2020Conditional Gaussian Distribution Learning for Open Set Recognitioncited by 104
  17. 2017Multi-Level Variational Autoencoder: Learning Disentangled Representations from Grouped Observationscited by 100
  18. 2019Are Disentangled Representations Helpful for Abstract Visual Reasoning?cited by 100
  19. 2019On the Transfer of Inductive Bias from Simulation to the Real World: a New Disentanglement Datasetcited by 100
  20. 2019When Explanations Lie: Why Many Modified BP Attributions Failcited by 92
  21. 2020ProtoPShare: Prototype Sharing for Interpretable Image Classification and Similarity Discoverycited by 89
  22. 2019Weakly Supervised Disentanglement with Guaranteescited by 82
  23. 2020Measuring Disentanglement: A Review of Metricscited by 70
  24. 2021This Looks Like That... Does it? Shortcomings of Latent Space Prototype Interpretability in Deep Networkscited by 65
  25. 2018Structured Disentangled Representationscited by 62
  26. 2020This Looks Like That, Because ... Explaining Prototypes for Interpretable Image Recognitioncited by 50
  27. 2019Interpretability Beyond Classification Output: Semantic Bottleneck Networkscited by 49
  28. 2021Symbols as a Lingua Franca for Bridging Human-AI Chasm for Explainable and Advisable AI Systemscited by 36
  29. 2021An Image is Worth More Than a Thousand Words: Towards Disentanglement in the Wildcited by 30
  30. 2021Interactive Disentanglement: Learning Concepts by Interacting with their Prototype Representationscited by 24
  31. 2019Weakly Supervised Disentanglement by Pairwise Similaritiescited by 22
  32. 2021On Causally Disentangled Representationscited by 22
  33. 2021Is Disentanglement all you need? Comparing Concept-based & Disentanglement Approachescited by 17
  34. 2021Local Disentanglement in Variational Auto-Encoders Using Jacobian L1 Regularizationcited by 17
  35. 2021Benchmarks, Algorithms, and Metrics for Hierarchical Disentanglementcited by 12
  36. 2019Variational Learning with Disentanglement-PyTorchcited by 5
  37. 2016Variational Autoencoders for Semi-supervised Text Classificationcited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2022: 1; 2023: 16; 2024: 27; 2025: 21; 2026: 8.

73 papers.

Sort:

By area:

  1. 2023Probabilistic Concept Bottleneck Modelscited by 101
  2. 2023Concept-based Explainable Artificial Intelligence: A Surveycited by 98
  3. 2023A Closer Look at the Intervention Procedure of Concept Bottleneck Modelscited by 64
  4. 2023Not All Neuro-Symbolic Concepts Are Created Equal: Analysis and Mitigation of Reasoning Shortcutscited by 63
  5. 2023Multi-dimensional concept discovery (MCD): A unifying framework with completeness guaranteescited by 50
  6. 2024Incremental Residual Concept Bottleneck Modelscited by 43
  7. 2023Learning to Receive Help: Intervention-Aware Concept Embedding Modelscited by 41
  8. 2024BEARS Make Neuro-Symbolic Models Aware of their Reasoning Shortcutscited by 37
  9. 2023Neuro-Symbolic Continual Learning: Knowledge, Reasoning Shortcuts and Concept Rehearsalcited by 33
  10. 2023Learning to Intervene on Concept Bottleneckscited by 29
  11. 2024Beyond Concept Bottleneck Models: How to Make Black Boxes Intervenable?cited by 28
  12. 2024Interpretable Concept Bottlenecks to Align Reinforcement Learning Agentscited by 23
38 more
  1. 2023Interpretability is in the Mind of the Beholder: A Causal Framework for Human-interpretable Representation Learningcited by 22
  2. 2024A Neuro-Symbolic Benchmark Suite for Concept Quality and Reasoning Shortcutscited by 19
  3. 2024Concept-Based Explanations in Computer Vision: Where Are We and Where Could We Go?cited by 18
  4. 2025A Comprehensive Survey on Self-Interpretable Neural Networkscited by 18
  5. 2024Concept Bottleneck Models Without Predefined Conceptscited by 17
  6. 2024Interpretable Concept-Based Memory Reasoningcited by 17
  7. 2024Reconsidering Faithfulness in Regular, Self-Explainable and Domain Invariant GNNscited by 16
  8. 2023Relational Concept Bottleneck Modelscited by 15
  9. 2025If Concept Bottlenecks are the Question, are Foundation Models the Answer?cited by 15
  10. 2024Neural Concept Bindercited by 14
  11. 2024Understanding Inter-Concept Relationships in Concept-Based Modelscited by 13
  12. 2025Shortcuts and Identifiability in Concept-based Models from a Neuro-Symbolic Lenscited by 13
  13. 2025Leakage and Interpretability in Concept-Based Modelscited by 13
  14. 2025Explain Yourself, Briefly! Self-Explaining Neural Networks with Concise Sufficient Reasonscited by 12
  15. 2023Interpretable and intervenable ultrasonography-based machine learning models for pediatric appendicitiscited by 11
  16. 2025Interpretable Generative Models through Post-hoc Concept Bottleneckscited by 11
  17. 2024Do Concept Bottleneck Models Respect Localities?cited by 10
  18. 2025Avoiding Leakage Poisoning: Concept Interventions Under Distribution Shiftscited by 9
  19. 2024A survey on Concept-based Approaches For Model Improvementcited by 8
  20. 2024Tree-Based Leakage Inspection and Control in Concept Bottleneck Modelscited by 8
  21. 2025There Was Never a Bottleneck in Concept Bottleneck Modelscited by 8
  22. 2024Counterfactual Concept Bottleneck Modelscited by 7
  23. 2023Selective Concept Models: Permitting Stakeholder Customisation at Test-Timecited by 6
  24. 2024Improving Intervention Efficacy via Concept Realignment in Concept Bottleneck Modelscited by 6
  25. 2025Neural Probabilistic Circuits: Enabling Compositional and Interpretable Predictions through Logical Reasoningcited by 6
  26. 2025Sample-efficient Learning of Concepts with Theoretical Guarantees: from Data to Concepts without Interventionscited by 6
  27. 2025SUB: Benchmarking CBM Generalization via Synthetic Attribute Substitutionscited by 6
  28. 2022Bort: Towards Explainable Neural Networks with Bounded Orthogonal Constraintcited by 5
  29. 2023How Faithful are Self-Explainable GNNs?cited by 5
  30. 2025LogicCBMs: Logic-Enhanced Concept-Based Learningcited by 5
  31. 2024Can we Constrain Concept Bottleneck Models to Learn Semantically Meaningful Input Features?cited by 4
  32. 2024Understanding Multimodal Deep Neural Networks: A Concept Selection Viewcited by 4
  33. 2025A Comprehensive Survey on the Risks and Limitations of Concept-based Modelscited by 4
  34. 2024Incorporating Expert Rules into Neural Networks in the Framework of Concept-Based Learningcited by 3
  35. 2024Linearly-Interpretable Concept Embedding Models for Text Analysiscited by 3
  36. 2025V-CEM: Bridging Performance and Intervenability in Concept-based Modelscited by 3
  37. 2025Towards Reasonable Concept Bottleneck Modelscited by 3
  38. 2025Interpretable Hierarchical Concept Reasoning through Attention-Guided Graph Learningcited by 3

50 most cited of 73 All 73 (JSON)