mlsanityData to 7 Oct 2026

Finding and Fixing Spurious Patterns with Explanations

arXiv 2106.02112 ↗First posted 3 Jun 2021cs.LGcited by 38

Not on the map. Why?

Abstract (arXiv)

Image classifiers often use spurious patterns, such as "relying on the presence of a person to detect a tennis racket, which do not generalize. In this work, we present an end-to-end pipeline for identifying and mitigating spurious patterns for such models, under the assumption that we have access to pixel-wise object-annotations. We start by identifying patterns such as "the model's prediction for tennis racket changes 63% of the time if we hide the people." Then, if a pattern is spurious, we mitigate it via a novel form of data augmentation. We demonstrate that our method identifies a diverse set of spurious patterns and that it mitigates them by producing a model that is both more accurate on a distribution where the spurious pattern is not helpful and more robust to distribution shift.

Key facts. Cited by 38 papers in the map (data through 7 Oct 2026). Builds on 40 of its 47 references that are in the map. First posted on arXiv 3 Jun 2021.

Builds on

40 of 47 references.

References · 40

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  3. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  4. 2014Microsoft COCO: Common Objects in Contextcited by 17,498
  5. 2016Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localizationcited by 6,690
  6. 2016"Why Should I Trust You?": Explaining the Predictions of Any Classifiercited by 5,040
  7. 2016Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answeringcited by 2,498
  8. 2013Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Mapscited by 2,348
  9. 2017Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)cited by 1,309
  10. 2018Sanity Checks for Saliency Mapscited by 1,149
  11. 2020Concept Bottleneck Modelscited by 1,022
  12. 2019Skin Lesion Analysis Toward Melanoma Detection 2018: A Challenge Hosted by the International Skin Imaging Collaboration (ISIC)cited by 657
28 more
  1. 2020Measuring Robustness to Natural Distribution Shifts in Image Classificationcited by 452
  2. 2020The Pitfalls of Simplicity Bias in Neural Networkscited by 349
  3. 2020Noise or Signal: The Role of Image Backgrounds in Object Recognitioncited by 342
  4. 2019Counterfactual Visual Explanationscited by 312
  5. 2017Right for the Right Reasons: Training Differentiable Models by Constraining their Explanationscited by 293
  6. 2018Explanations based on the Missing: Towards Contrastive Explanations with Pertinent Negativescited by 261
  7. 2018Women also Snowboard: Overcoming Bias in Captioning Modelscited by 236
  8. 2019EdgeConnect: Generative Image Inpainting with Adversarial Edge Learningcited by 161
  9. 2020Counterfactual Samples Synthesizing for Robust Visual Question Answeringcited by 153
  10. 2020Understanding the Failure Modes of Out-of-Distribution Generalizationcited by 146
  11. 2020Debugging Tests for Model Explanationscited by 146
  12. 2019Interpretations are useful: penalizing explanations to align neural networks with prior knowledgecited by 144
  13. 2020REVISE: A Tool for Measuring and Mitigating Bias in Visual Datasetscited by 136
  14. 2021Counterfactual Generative Networkscited by 111
  15. 2019Towards Causal VQA: Revealing and Reducing Spurious Correlations by Invariant and Covariant Semantic Editingcited by 108
  16. 2020Don't Judge an Object by Its Context: Learning to Overcome Contextual Biascited by 91
  17. 2020Model Patching: Closing the Subgroup Performance Gap with Data Augmentationcited by 82
  18. 2020Self-training Avoids Using Spurious Features Under Domain Shiftcited by 69
  19. 2018Hierarchical interpretations for neural network predictionscited by 66
  20. 2020Learning What Makes a Difference from Counterfactual Examples and Gradient Supervisioncited by 66
  21. 2019SpatialSense: An Adversarially Crowdsourced Benchmark for Spatial Relation Recognitioncited by 55
  22. 2018Not Using the Car to See the Sidewalk: Quantifying and Controlling the Effects of Context in Classification and Segmentationcited by 49
  23. 2017Weakly-supervised learning of visual relationscited by 36
  24. 2020Gender Slopes: Counterfactual Fairness for Computer Vision Models by Attribute Manipulationcited by 36
  25. 2020Towards Accuracy-Fairness Paradox: Adversarial Example-based Data Augmentation for Visual Debiasingcited by 32
  26. 2020Explaining Groups of Points in Low-Dimensional Representationscited by 16
  27. 2020Causality-aware counterfactual confounding adjustment for feature representations learned by deep modelscited by 2
  28. 2020Interventional Domain Adaptationcited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 1; 2022: 7; 2023: 14; 2024: 9; 2025: 4; 2026: 3.

38 papers.

Sort:

By area:

  1. 2021Salient ImageNet: How to discover spurious features in Deep Learning?cited by 114
  2. 2023Discover and Cure: Concept-aware Mitigation of Spurious Correlationcited by 71
  3. 2022Post hoc Explanations may be Ineffective for Detecting Unknown Spurious Correlationcited by 59
  4. 2022Spurious Features Everywhere -- Large-Scale Detection of Harmful Spurious Features in ImageNetcited by 41
  5. 2024A Sober Look at the Robustness of CLIPs to Spurious Featurescited by 38
  6. 2024MM-SpuBench: Towards Better Understanding of Spurious Biases in Multimodal LLMscited by 33
  7. 2023Towards Trustworthy and Aligned Machine Learning: A Data-centric Survey with Causality Perspectivescited by 30
  8. 2023Balancing the Picture: Debiasing Vision-Language Datasets with Synthetic Contrast Setscited by 27
  9. 2023Seeing is not Believing: Robust Reinforcement Learning against Spurious Correlationcited by 25
  10. 2024Navigating Shortcuts, Spurious Correlations, and Confounders: From Origins via Detection to Mitigationcited by 22
  11. 2023The Impact of Imperfect XAI on Human-AI Decision-Makingcited by 21
  12. 2023From Hope to Safety: Unlearning Biases of Deep Models via Gradient Penalization in Latent Spacecited by 20
26 more
  1. 2023CLIP-QDA: An Explainable Concept Bottleneck Modelcited by 13
  2. 2023How Spurious Features Are Memorized: Precise Analysis for Random and NTK Featurescited by 11
  3. 2023DiG-IN: Diffusion Guidance for Investigating Networks -- Uncovering Classifier Differences Neuron Visualisations and Visual Counterfactual Explanationscited by 11
  4. 2022Identifying Spurious Correlations and Correcting them with an Explanation-based Learningcited by 10
  5. 2022Invariant Learning via Diffusion Dreamed Distribution Shiftscited by 10
  6. 2022AGRO: Adversarial Discovery of Error-prone groups for Robust Optimizationcited by 10
  7. 2024Learning Robust Classifiers with Self-Guided Spurious Correlation Mitigationcited by 10
  8. 2023Preemptively Pruning Clever-Hans Strategies in Deep Neural Networkscited by 9
  9. 2023PRIME: Prioritizing Interpretability in Failure Mode Extractioncited by 9
  10. 2024Sharpness-Aware Minimization Enhances Feature Quality via Balanced Learningcited by 9
  11. 2025Spurious Correlations in High Dimensional Regression: The Roles of Regularization, Simplicity Bias and Over-Parameterizationcited by 9
  12. 2022Perspectives on Incorporating Expert Feedback into Model Updatescited by 7
  13. 2024Spuriousness-Aware Meta-Learning for Learning Robust Classifierscited by 7
  14. 2024Visual Data Diagnosis and Debiasing with Concept Graphscited by 6
  15. 2022Towards a More Rigorous Science of Blindspot Discovery in Image Classification Modelscited by 5
  16. 2026Spurious Correlation Learning in Preference Optimization: Mechanisms, Consequences, and Mitigation via Tie Trainingcited by 4
  17. 2025BiMa: Towards Biases Mitigation for Text-Video Retrieval via Scene Element Guidancecited by 2
  18. 2023Addressing Weak Decision Boundaries in Image Classification by Leveraging Web Search and Generative Modelscited by 1
  19. 2023Detecting Spurious Correlations via Robust Visual Concepts in Real and AI-Generated Image Classificationcited by 1
  20. 2023MaskSearch: Querying Image Masks at Scalecited by 0
  21. 2024Demonstration of MaskSearch: Efficiently Querying Image Masks for Machine Learning Workflowscited by 0
  22. 2024Effective Guidance for Model Attention with Simple Yes-no Annotationscited by 0
  23. 2025Token-Based Detection of Spurious Correlations in Vision Transformerscited by 0
  24. 2025Human Cognitive Biases in Explanation-Based Interaction: The Case of Within and Between Session Order Effectcited by 0
  25. 2026Mitigating Shortcut Learning via Feature Disentanglement in Medical Imaging: A Benchmark Studycited by 0
  26. 2026"World Knowledge" in the Weights: Reading Concept Circuits of Vision Transformerscited by 0