mlsanityData to 7 Oct 2026

CLIP-Dissect: Automatic Description of Neuron Representations in Deep Vision Networks

arXiv 2204.10965 ↗First posted 23 Apr 2022cs.CV cs.AI cs.LGcited by 144

Not on the map. Why?

Abstract (arXiv)

In this paper, we propose CLIP-Dissect, a new technique to automatically describe the function of individual hidden neurons inside vision networks. CLIP-Dissect leverages recent advances in multimodal vision/language models to label internal neurons with open-ended concepts without the need for any labeled data or human examples. We show that CLIP-Dissect provides more accurate descriptions than existing methods for last layer neurons where the ground-truth is available as well as qualitatively good descriptions for hidden layer neurons. In addition, our method is very flexible: it is model agnostic, can easily handle new concepts and can be extended to take advantage of better multimodal models in the future. Finally CLIP-Dissect is computationally efficient and can label all neurons from five layers of ResNet-50 in just 4 minutes, which is more than 10 times faster than existing methods. Our code is available at https://github.com/Trustworthy-ML-Lab/CLIP-dissect. Finally, crowdsourced user study results are available at Appendix B to further support the effectiveness of our method.

Key facts. Cited by 144 papers in the map (data through 7 Oct 2026). Builds on 13 of its 26 references that are in the map. First posted on arXiv 23 Apr 2022.

Builds on

13 of 26 references.

References · 13

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  3. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  4. 2013Visualizing and Understanding Convolutional Networkscited by 2,278
  5. 2021Florence: A New Foundation Model for Computer Visioncited by 802
  6. 2017Network Dissection: Quantifying Interpretability of Deep Visual Representationscited by 703
  7. 2021LiT: Zero-Shot Transfer with Locked-image text Tuningcited by 572
  8. 2020Understanding the Role of Individual Units in a Deep Neural Networkcited by 254
  9. 2021Combined Scaling for Zero-shot Transfer Learningcited by 182
  10. 2014Object Detectors Emerge in Deep Scene CNNscited by 178
  11. 2020Compositional Explanations of Neuronscited by 172
  12. 2022Natural Language Descriptions of Deep Visual Featurescited by 136
1 more
  1. 2020Towards Unique and Informative Captioning of Imagescited by 25

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 1; 2023: 19; 2024: 34; 2025: 46; 2026: 44.

144 papers.

Sort:

By area:

  1. 2023Label-Free Concept Bottleneck Modelscited by 282
  2. 2022Toward Transparent AI: A Survey on Interpreting the Inner Structures of Deep Neural Networkscited by 126
  3. 2023Concept-based Explainable Artificial Intelligence: A Surveycited by 97
  4. 2024Discover-then-Name: Task-Agnostic Concept Bottlenecks via Automated Concept Discoverycited by 91
  5. 2025Sparse Autoencoders Learn Monosemantic Features in Vision-Language Modelscited by 70
  6. 2024VLG-CBM: Training Concept Bottleneck Models with Vision-Language Guidancecited by 65
  7. 2023Text-To-Concept (and Back) via Cross-Model Alignmentcited by 63
  8. 2024A Multimodal Automated Interpretability Agentcited by 62
  9. 2025Interpreting CLIP with Hierarchical Sparse Autoencoderscited by 54
  10. 2024Explainable and Interpretable Multimodal Large Language Models: A Comprehensive Surveycited by 52
  11. 2024Concept Bottleneck Large Language Modelscited by 41
  12. 2024Interpreting the Second-Order Effects of Neurons in CLIPcited by 36
38 more
  1. 2025A Survey on Mechanistic Interpretability for Multi-Modal Foundation Modelscited by 34
  2. 2023Identifying Interpretable Subspaces in Image Representationscited by 32
  3. 2023FIND: A Function Description Benchmark for Evaluating Interpretability Methodscited by 32
  4. 2023Coarse-to-Fine Concept Bottleneck Modelscited by 31
  5. 2023Labeling Neural Representations with Inverse Recognitioncited by 31
  6. 2024MMNeuron: Discovering Neuron-Level Domain-Specific Interpretation in Multimodal Large Language Modelcited by 26
  7. 2025From Colors to Classes: Emergence of Concepts in Vision Transformerscited by 23
  8. 2024Decomposing and Editing Predictions by Modeling Model Computationcited by 19
  9. 2024Linear Explanations for Individual Neuronscited by 19
  10. 2024Decomposing and Interpreting Image Representations via Text in ViTs Beyond CLIPcited by 19
  11. 2024CoSy: Evaluating Textual Explanations of Neuronscited by 17
  12. 2024Learning Discrete Concepts in Latent Hierarchical Modelscited by 17
  13. 2024Algorithmic Capabilities of Random Transformerscited by 17
  14. 2024DiffusionPID: Interpreting Diffusion via Partial Information Decompositioncited by 16
  15. 2023Towards a fuller understanding of neurons with Clustered Compositional Explanationscited by 15
  16. 2024WWW: A Unified Framework for Explaining What, Where and Why of Neural Networks by Interpretation of Neuron Conceptscited by 14
  17. 2023From Neural Activations to Concepts: A Survey on Explaining Concepts in Neural Networkscited by 13
  18. 2023The Importance of Prompt Tuning for Automated Neuron Explanationscited by 12
  19. 2025Show and Tell: Visually Explainable Deep Neural Nets via Spatially-Aware Concept Bottleneck Modelscited by 12
  20. 2024Crafting Interpretable Embeddings by Asking LLMs Questionscited by 11
  21. 2025Interpretable Generative Models through Post-hoc Concept Bottleneckscited by 11
  22. 2024Interpreting Neurons in Deep Vision Networks with Language Modelscited by 10
  23. 2024LG-CAV: Train Any Concept Activation Vector with Language Guidancecited by 9
  24. 2024Adaptive Concept Bottleneck for Foundation Models Under Distribution Shiftscited by 9
  25. 2025V2C-CBM: Building Concept Bottlenecks with Vision-to-Concept Tokenizercited by 9
  26. 2025CoE: Chain-of-Explanation via Automatic Visual Concept Circuit Description and Polysemanticity Quantificationcited by 9
  27. 2023DISCOVER: Making Vision Networks Interpretable via Competition and Dissectioncited by 8
  28. 2025Effective Skill Unlearning through Intervention and Abstentioncited by 8
  29. 2023Adversarial Attacks on the Interpretation of Neuron Activation Maximizationcited by 7
  30. 2024AND: Audio Network Dissection for Interpreting Deep Acoustic Modelscited by 7
  31. 2024Crafting Large Language Models for Enhanced Interpretabilitycited by 7
  32. 2024Abstraction Alignment: Comparing Model-Learned and Human-Encoded Conceptual Relationshipscited by 7
  33. 2024MINER: Mining the Underlying Pattern of Modality-Specific Neurons in Multimodal Large Language Modelscited by 7
  34. 2025Discovering Influential Neuron Path in Vision Transformerscited by 7
  35. 2023Interpreting CLIP: Insights on the Robustness to ImageNet Distribution Shiftscited by 6
  36. 2024On the Value of Labeled Data and Symbolic Methods for Hidden Neuron Activation Analysiscited by 6
  37. 2024LLM-assisted Concept Discovery: Automatically Identifying and Explaining Neuron Functionscited by 6
  38. 2025Mysteries of the Deep: Role of Intermediate Representations in Out of Distribution Detectioncited by 6

50 most cited of 144 All 144 (JSON)