mlsanityData to 7 Oct 2026

Shortcut Mitigation via Spurious-Positive Samples

arXiv 2605.13340 ↗First posted 13 May 2026cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

Shortcut mitigation strategies commonly rely on training data annotations, group-balanced held-out data or the presence of all groups, i.e., all combinations of (spurious) attributes and classes, in the training data. However, these requirements are rarely met in practice. We instead propose a method for targeted model analysis to identify a small set of instances in which the model relies on spurious attributes. Using that set and following ``this feature should not be used for prediction'' reasoning, we identify highly relevant neurons in an intermediate layer and regularize their impact. This ensures that models learn to depend on informative features rather than being right for the wrong reasons, thereby improving robustness without requiring additional balanced held-out data or annotations.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 28 of its 37 references that are in the map. First posted on arXiv 13 May 2026.

Builds on

28 of 37 references.

References · 28

  1. 2014Deep Learning Face Attributes in the Wildcited by 3,679
  2. 2018UMAP: Uniform Manifold Approximation and Projection for Dimension Reductioncited by 2,841
  3. 2018ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustnesscited by 1,430
  4. 2019Skin Lesion Analysis Toward Melanoma Detection 2018: A Challenge Hosted by the International Skin Imaging Collaboration (ISIC)cited by 657
  5. 2021Just Train Twice: Improving Group Robustness without Training Group Informationcited by 522
  6. 2022Last Layer Re-Training is Sufficient for Robustness to Spurious Correlationscited by 379
  7. 2018Confounding variables can degrade generalization performance of radiological deep learning modelscited by 374
  8. 2020An Investigation of Why Overparameterization Exacerbates Spurious Correlationscited by 310
  9. 2017Right for the Right Reasons: Training Differentiable Models by Constraining their Explanationscited by 293
  10. 2022Correct-N-Contrast: A Contrastive Approach for Improving Robustness to Spurious Correlationscited by 171
  11. 2019Interpretations are useful: penalizing explanations to align neural networks with prior knowledgecited by 144
  12. 2024AttnLRP: Attention-Aware Layer-Wise Relevance Propagation for Transformerscited by 116
16 more
  1. 2021Can Subnetwork Structure be the Key to Out-of-Distribution Generalization?cited by 75
  2. 2023Discover and Cure: Concept-aware Mitigation of Spurious Correlationcited by 71
  3. 2023Simple and Fast Group Robustness by Automatic Feature Reweightingcited by 68
  4. 2022MaskTune: Mitigating Spurious Correlations by Forcing to Explorecited by 66
  5. 2020Robustness to Spurious Correlations via Human Annotationscited by 62
  6. 2023Towards Last-layer Retraining for Group Robustness with Fewer Annotationscited by 54
  7. 2022Optimizing Relevance Maps of Vision Transformers Improves Robustnesscited by 41
  8. 2022On Guiding Visual Attention with Language Specificationcited by 30
  9. 2024PURE: Turning Polysemantic Neurons Into Pure Features by Identifying Relevant Circuitscited by 29
  10. 2022Training Debiased Subnetworks with Contrastive Weight Pruningcited by 16
  11. 2024Out of Spuriousity: Improving Robustness to Spurious Correlations without Group Annotationscited by 9
  12. 2024Efficient Bias Mitigation Without Privileged Informationcited by 8
  13. 2023Using Early Readouts to Mediate Featural Bias in Distillationcited by 7
  14. 2024Visual Data Diagnosis and Debiasing with Concept Graphscited by 6
  15. 2025Efficient Unsupervised Shortcut Learning Detection and Mitigation in Transformerscited by 6
  16. 2025An XAI-based Analysis of Shortcut Learning in Neural Networkscited by 2

Built on by

Not cited yet.