mlsanityData to 7 Oct 2026

Now You See Me (CME): Concept-based Model Extraction

arXiv 2010.13233 ↗First posted 25 Oct 2020cs.LGcited by 51

Not on the map. Why?

Abstract (arXiv)

Deep Neural Networks (DNNs) have achieved remarkable performance on a range of tasks. A key step to further empowering DNN-based approaches is improving their explainability. In this work we present CME: a concept-based model extraction framework, used for analysing DNN models via concept-based extracted models. Using two case studies (dSprites, and Caltech UCSD Birds), we demonstrate how CME can be used to (i) analyse the concept information learned by a DNN model (ii) analyse how a DNN uses this concept information when predicting output labels (iii) identify key concept information that can further improve DNN predictive performance (for one of the case studies, we showed how model accuracy can be improved by over 14%, using only 30% of the available concepts).

Key facts. Cited by 51 papers in the map (data through 7 Oct 2026). First posted on arXiv 25 Oct 2020.

Builds on

No reference list yet.

Built on by

Cited by year

Citing papers per year: 2021: 7; 2022: 9; 2023: 13; 2024: 15; 2025: 7.

51 papers.

Sort:

By area:

  1. 2023Concept-based Explainable Artificial Intelligence: A Surveycited by 97
  2. 2022Post hoc Explanations may be Ineffective for Detecting Unknown Spurious Correlationcited by 59
  3. 2023Interpretable Neural-Symbolic Concept Reasoningcited by 56
  4. 2021Entropy-based Logic Explanations of Neural Networkscited by 51
  5. 2021Logic Explained Networkscited by 45
  6. 2021A Framework for Learning Ante-hoc Explainable Models via Conceptscited by 45
  7. 2024Incremental Residual Concept Bottleneck Modelscited by 43
  8. 2023Learning to Receive Help: Intervention-Aware Concept Embedding Modelscited by 41
  9. 2024Energy-Based Concept Bottleneck Models: Unifying Prediction, Concept Intervention, and Probabilistic Interpretationscited by 33
  10. 2023Towards Robust Metrics for Concept Representation Evaluationcited by 29
  11. 2023Coherent Concept-based Explanations in Medical Image and Its Application to Skin Lesion Diagnosiscited by 26
  12. 2023Interpretability is in the Mind of the Beholder: A Causal Framework for Human-interpretable Representation Learningcited by 22
38 more
  1. 2022Concept Embedding Analysis: A Reviewcited by 20
  2. 2022Encoding Concepts in Graph Neural Networkscited by 18
  3. 2021Is Disentanglement all you need? Comparing Concept-based & Disentanglement Approachescited by 17
  4. 2024Concept-Based Explanations in Computer Vision: Where Are We and Where Could We Go?cited by 17
  5. 2022When are Post-hoc Conceptual Explanations Identifiable?cited by 14
  6. 2024Understanding Inter-Concept Relationships in Concept-Based Modelscited by 13
  7. 2021Algorithmic Concept-based Explainable Reasoningcited by 12
  8. 2023Evaluating the Stability of Semantic Concept Representations in CNNs for Robust Explainabilitycited by 11
  9. 2024Concept-Attention Whitening for Interpretable Skin Lesion Diagnosiscited by 11
  10. 2024Do Concept Bottleneck Models Respect Localities?cited by 10
  11. 2022Provable concept learning for interpretable predictions using variational autoencoderscited by 9
  12. 2023Concept Distillation: Leveraging Human-Centered Explanations for Model Improvementcited by 9
  13. 2025Avoiding Leakage Poisoning: Concept Interventions Under Distribution Shiftscited by 9
  14. 2024A survey on Concept-based Approaches For Model Improvementcited by 8
  15. 2023Concept-Centric Transformers: Enhancing Model Interpretability through Object-Centric Concept Learning within a Shared Global Workspacecited by 7
  16. 2021Failing Conceptually: Concept-Based Explanations of Dataset Shiftcited by 6
  17. 2022Towards learning to explain with concept bottleneck models: mitigating information leakagecited by 6
  18. 2025Neural Probabilistic Circuits: Enabling Compositional and Interpretable Predictions through Logical Reasoningcited by 6
  19. 2021PyTorch, Explain! A Python library for Logic Explained Networkscited by 5
  20. 2022Towards Human-Interpretable Prototypes for Visual Assessment of Image Classification Modelscited by 4
  21. 2023Revealing Similar Semantics Inside CNNs: An Interpretable Concept-based Comparison of Feature Spacescited by 4
  22. 2023Statistically Significant Concept-based Explanation of Image Classifiers via Model Knockoffscited by 4
  23. 2023LR-XFL: Logical Reasoning-based Explainable Federated Learningcited by 4
  24. 2024Understanding Multimodal Deep Neural Networks: A Concept Selection Viewcited by 4
  25. 2024Concept-based Explainable Malignancy Scoring on Pulmonary Nodules in CT Imagescited by 2
  26. 2024TAB: Transformer Attention Bottlenecks enable User Intervention and Debugging in Vision-Language Modelscited by 2
  27. 2022The Influence of Explainable Artificial Intelligence: Nudging Behaviour or Boosting Capability?cited by 1
  28. 2024ECATS: Explainable-by-design concept-based anomaly detection for time seriescited by 1
  29. 2025BarkXAI: A Lightweight Post-Hoc Explainable Method for Tree Species Classification with Quantifiable Conceptscited by 1
  30. 2025Walking the Web of Concept-Class Relationships in Incrementally Trained Interpretable Modelscited by 1
  31. 2025Understanding and Improving Adversarial Robustness of Neural Probabilistic Circuitscited by 1
  32. 2025Flexible Concept Bottleneck Modelcited by 1
  33. 2022Explainer Divergence Scores (EDS): Some Post-Hoc Explanations May be Effective for Detecting Unknown Spurious Correlationscited by 0
  34. 2023GCI: A (G)raph (C)oncept (I)nterpretation Frameworkcited by 0
  35. 2024Learning Decomposable and Debiased Representations via Attribute-Centric Information Bottleneckscited by 0
  36. 2024Enhancing Interpretability of Vertebrae Fracture Grading using Human-interpretable Prototypescited by 0
  37. 2024MulCPred: Learning Multi-modal Concepts for Explainable Pedestrian Action Predictioncited by 0
  38. 2024Are Explanations Helpful? A Comparative Analysis of Explainability Methods in Skin Lesion Classifierscited by 0

50 most cited of 51 All 51 (JSON)