mlsanityData to 7 Oct 2026

Enhancing Concept Localization in CLIP-based Concept Bottleneck Models

arXiv 2510.07115 ↗First posted 8 Oct 2025cs.CVcited by 6

Not on the map. Why?

Abstract (arXiv)

This paper addresses explainable AI (XAI) through the lens of Concept Bottleneck Models (CBMs) that do not require explicit concept annotations, relying instead on concepts extracted using CLIP in a zero-shot manner. We show that CLIP, which is central in these techniques, is prone to concept hallucination, incorrectly predicting the presence or absence of concepts within an image in scenarios used in numerous CBMs, hence undermining the faithfulness of explanations. To mitigate this issue, we introduce Concept Hallucination Inhibition via Localized Interpretability (CHILI), a technique that disentangles image embeddings and localizes pixels corresponding to target concepts. Furthermore, our approach supports the generation of saliency-based explanations that are more interpretable.

Key facts. Cited by 6 papers in the map (data through 7 Oct 2026). Builds on 59 of its 70 references that are in the map. First posted on arXiv 8 Oct 2025.

Builds on

59 of 70 references.

References · 59

  1. 2016Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localizationcited by 6,690
  2. 2017A Unified Approach to Interpreting Model Predictionscited by 6,488
  3. 2016"Why Should I Trust You?": Explaining the Predictions of Any Classifiercited by 5,040
  4. 2023Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detectioncited by 3,274
  5. 2015Learning Deep Features for Discriminative Localizationcited by 2,430
  6. 2022Conditional Prompt Learning for Vision-Language Modelscited by 1,427
  7. 2020Concept Bottleneck Modelscited by 1,022
  8. 2019Analyzing Multi-Head Self-Attention: Specialized Heads Do the Heavy Lifting, the Rest Can Be Prunedcited by 844
  9. 2023Vision Transformers Need Registerscited by 844
  10. 2018RISE: Randomized Input Sampling for Explanation of Black-box Modelscited by 825
  11. 2020Quantifying Attention Flow in Transformerscited by 759
  12. 2021Multimodal datasets: misogyny, pornography, and malignant stereotypescited by 326
38 more
  1. 2021Generic Attention-model Explainability for Interpreting Bi-Modal and Encoder-Decoder Transformerscited by 316
  2. 2022Language in a Bottle: Language Model Guided Concept Bottlenecks for Interpretable Image Classificationcited by 310
  3. 2023Label-Free Concept Bottleneck Modelscited by 282
  4. 2016Layer-wise Relevance Propagation for Neural Networks with Local Renormalization Layerscited by 197
  5. 2019Full-Gradient Representation for Neural Network Visualizationcited by 189
  6. 2023Interpreting CLIP's Image Representation via Text-Based Decompositioncited by 189
  7. 2024Interpreting CLIP with Sparse Linear Concept Embeddings (SpLiCE)cited by 116
  8. 2020Use HiResCAM instead of Grad-CAM for faithful explanations of convolutional neural networkscited by 79
  9. 2023Learning Concise and Descriptive Attributes for Visual Recognitioncited by 73
  10. 2024VLG-CBM: Training Concept Bottleneck Models with Vision-Language Guidancecited by 65
  11. 2023Text-To-Concept (and Back) via Cross-Model Alignmentcited by 63
  12. 2024A Multimodal Automated Interpretability Agentcited by 62
  13. 2024Two Effects, One Trigger: On the Modality Gap, Object Bias, and Information Imbalance in Contrastive Vision-Language Modelscited by 55
  14. 2024Stochastic Concept Bottleneck Modelscited by 49
  15. 2023Robust and Interpretable Medical Image Classifiers via Concept Bottleneck Modelscited by 40
  16. 2023Vamos: Versatile Action Models for Video Understandingcited by 40
  17. 2024CLIP the Bias: How Useful is Balancing Data in Multimodal Learning?cited by 35
  18. 2023Identifying Interpretable Subspaces in Image Representationscited by 32
  19. 2023Vision-Language Models Performing Zero-Shot Tasks Exhibit Gender-based Disparitiescited by 31
  20. 2023Coarse-to-Fine Concept Bottleneck Modelscited by 31
  21. 2021EXplainable Neural-Symbolic Learning (X-NeSyL) methodology to fuse deep learning representations with expert knowledge graphs: the MonuMAI cultural heritage use casecited by 30
  22. 2023A ChatGPT Aided Explainable Framework for Zero-Shot Medical Image Diagnosiscited by 30
  23. 2023STAIR: Learning Sparse Text and Image Representation in Grounded Tokenscited by 23
  24. 2023Sparse Linear Concept Discovery Modelscited by 22
  25. 2023ChatGPT-Powered Hierarchical Comparisons for Image Classificationcited by 22
  26. 2023Concept Bottleneck with Visual Concept Filtering for Explainable Medical Image Classificationcited by 21
  27. 2024XCoOp: Explainable Prompt Learning for Computer-Aided Diagnosis via Concept-guided Context Optimizationcited by 20
  28. 2024Evolving Interpretable Visual Classifiers with Large Language Modelscited by 18
  29. 2024BendVLM: Test-Time Debiasing of Vision-Language Embeddingscited by 17
  30. 2024Investigating and Mitigating Object Hallucinations in Pretrained Vision-Language (CLIP) Modelscited by 15
  31. 2025If Concept Bottlenecks are the Question, are Foundation Models the Answer?cited by 15
  32. 2022Spuriosity Rankings: Sorting Data to Measure and Mitigate Biasescited by 13
  33. 2023CLIP-QDA: An Explainable Concept Bottleneck Modelcited by 13
  34. 2022Greybox XAI: a Neural-Symbolic learning framework to produce interpretable predictions for image classificationcited by 11
  35. 2023ARTxAI: Explainable Artificial Intelligence Curates Deep Representation Learning for Artistic Images using Fuzzy Techniquescited by 10
  36. 2023Driving through the Concept Gridlock: Unraveling Explainability Bottlenecks in Automated Drivingcited by 9
  37. 2023Towards Concept-based Interpretability of Skin Lesion Diagnosis using Vision-Language Modelscited by 9
  38. 2023Q-SENN: Quantized Self-Explaining Neural Networkscited by 9

50 most cited of 59 All 59 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 6.

6 papers.

  1. 2026Disentangling Hallucinations: Orthogonal Semantic Projection for Robust Interpretabilitycited by 1
  2. 2026Towards Faithful Multimodal Concept Bottleneck Modelscited by 0
  3. 2026A Composite Activation Function for Learning Stable Binary Representationscited by 0
  4. 2026Concepts Worth Having: Refining VLM-Guided Concept Bottleneck Models with Minimal Annotationscited by 0
  5. 2026MorphoSHAP: Rethinking the Unit of Attribution in Explanation for Deep Visual Modelscited by 0
  6. 2026CHOQOLATE: Organizing Concept Bottleneck Latent Spaces with Choquet Integralscited by 0