mlsanityData to 14 Oct 2026

Probabilistic Stability Guarantees for Feature Attributions

arXiv 2504.13787 ↗First posted 18 Apr 2025cs.LG cs.AIcited by 9

Not on the map. Why?

Abstract (arXiv)

Stability guarantees have emerged as a principled way to evaluate feature attributions, but existing certification methods rely on heavily smoothed classifiers and often produce conservative guarantees. To address these limitations, we introduce soft stability and propose a simple, model-agnostic, sample-efficient stability certification algorithm (SCA) that yields non-trivial and interpretable guarantees for any attribution method. Moreover, we show that mild smoothing achieves a more favorable trade-off between accuracy and stability, avoiding the aggressive compromises made in prior certification methods. To explain this behavior, we use Boolean function analysis to derive a novel characterization of stability under smoothing. We evaluate SCA on vision and language tasks and demonstrate the effectiveness of soft stability in measuring the robustness of explanation methods.

Key facts. Cited by 9 papers in the map (data through 14 Oct 2026). Builds on 61 of its 77 references that are in the map. First posted on arXiv 18 Apr 2025.

Builds on

61 of 77 references.

References · 61

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  3. 2019RoBERTa: A Robustly Optimized BERT Pretraining Approachcited by 13,372
  4. 2017A Unified Approach to Interpreting Model Predictionscited by 6,555
  5. 2017Axiomatic Attribution for Deep Networkscited by 3,384
  6. 2016Categorical Reparameterization with Gumbel-Softmaxcited by 2,739
  7. 2013Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Mapscited by 2,371
  8. 2017Understanding Black-box Predictions via Influence Functionscited by 1,741
  9. 2019Certified Adversarial Robustness via Randomized Smoothingcited by 1,340
  10. 2017Explanation in Artificial Intelligence: Insights from the Social Sciencescited by 1,206
  11. 2018Sanity Checks for Saliency Mapscited by 1,156
  12. 2017SmoothGrad: removing noise by adding noisecited by 1,109
38 more
  1. 2019Fooling LIME and SHAP: Adversarial Attacks on Post hoc Explanation Methodscited by 443
  2. 2015Evaluating the visualization of what a Deep Neural Network has learnedcited by 419
  3. 2018Lipschitz regularity of deep neural networks: analysis and efficient estimationcited by 415
  4. 2019The many Shapley values for model explanationcited by 344
  5. 2017The (Un)reliability of saliency methodscited by 333
  6. 2019On the (In)fidelity and Sensitivity for Explanationscited by 310
  7. 2022From Anecdotal Evidence to Quantitative Evaluation Methods: A Systematic Review on Evaluating Explainable AIcited by 302
  8. 2019Efficient and Accurate Estimation of Lipschitz Constants for Deep Neural Networkscited by 293
  9. 2023Faithful Chain-of-Thought Reasoningcited by 288
  10. 2020Influence Functions in Deep Learning Are Fragilecited by 252
  11. 2018Sorting out Lipschitz function approximationcited by 213
  12. 2020A Diagnostic Study of Explainability Techniques for Text Classificationcited by 159
  13. 2022Towards Faithful Model Explanation in NLP: A Surveycited by 130
  14. 2022OpenXAI: Towards a Transparent Evaluation of Model Explanationscited by 126
  15. 2024Benchmarking LLMs via Uncertainty Quantificationcited by 126
  16. 2022Safe Planning in Dynamic Environments using Conformal Predictioncited by 119
  17. 2022A Consistent and Efficient Evaluation Strategy for Attribution Methodscited by 102
  18. 2021Do Feature Attribution Methods Correctly Attribute Features?cited by 98
  19. 2021A Comprehensive Taxonomy for Explainable Artificial Intelligence: A Systematic Survey of Surveys on Methods and Conceptscited by 91
  20. 2021The Out-of-Distribution Problem in Explainability and Search Methods for Feature Importance Explanationscited by 87
  21. 2022Impossibility Theorems for Feature Attributioncited by 78
  22. 2019Towards a Rigorous Evaluation of XAI Methods on Time Seriescited by 67
  23. 2022Post hoc Explanations may be Ineffective for Detecting Unknown Spurious Correlationcited by 59
  24. 2018Limitations of the Lipschitz constant as a defense against adversarial examplescited by 49
  25. 2019Interpretable and Fine-Grained Visual Explanations for Convolutional Neural Networkscited by 49
  26. 2022Towards Formal XAI: Formally Approximate Minimal Explanations of Neural Networkscited by 42
  27. 2021Improved, Deterministic Smoothing for L_1 Certified Robustnesscited by 39
  28. 2022Missingness Bias in Model Debuggingcited by 38
  29. 2022DIME: Fine-grained Interpretations of Multimodal Models via Disentangled Local Explanationscited by 34
  30. 2021Provably efficient, succinct, and precise explanationscited by 32
  31. 2022Explainable Deep Learning Methods in Medical Image Classification: A Surveycited by 32
  32. 2021Probabilistic Sufficient Explanationscited by 24
  33. 2023On the Robustness of Removal-Based Feature Attributionscited by 19
  34. 2020Explainable Machine Learning for Public Policy: Use Cases, Gaps, and Research Directionscited by 17
  35. 2022Chordal Sparsity for Lipschitz Constant Estimation of Deep Neural Networkscited by 15
  36. 2023Stability Guarantees for Feature Attributions with Multiplicative Smoothingcited by 14
  37. 2025On the Computational Tractability of the (Many) Shapley Valuescited by 12
  38. 2020Challenging common interpretability assumptions in feature attribution explanationscited by 11

50 most cited of 61 All 61 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 5; 2026: 4.

9 papers.

  1. 2025Explaining, Fast and Slow: Abstraction and Refinement of Provable Explanationscited by 13
  2. 2026Formal Mechanistic Interpretability: Automated Circuit Discovery with Provable Guaranteescited by 11
  3. 2025SHAP Meets Tensor Networks: Provably Tractable Explanations with Parallelismcited by 8
  4. 2026Interpretability Can Be Actionablecited by 7
  5. 2025Additive Models Explained: A Computational Complexity Approachcited by 3
  6. 2025Invisible Attributes, Visible Biases: Exploring Demographic Shortcuts in MRI-based Alzheimer's Disease Classificationcited by 2
  7. 2026Unifying Formal Explanations: A Complexity-Theoretic Perspectivecited by 2
  8. 2026Missingness Bias Calibration in Feature Attribution Explanationscited by 1
  9. 2025When LRP Diverges from Leave-One-Out in Transformerscited by 0