# Don't blame Dataset Shift! Shortcut Learning due to Gradients and Cross Entropy

arXiv: https://arxiv.org/abs/2308.12553 · HTML: https://mlsanity.com/paper/2308.12553/

## Key facts

- Cited by 30 papers in the map (data through 14 Oct 2026).
- Builds on 40 of its 57 references that are in the map.
- First posted on arXiv 24 Aug 2023.

## Builds on (most cited first, 30 of 40)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [Deep Learning Face Attributes in the Wild](https://mlsanity.com/paper/1411.7766/) (2014)
- [Shortcut Learning in Deep Neural Networks](https://mlsanity.com/paper/2004.07780/) (2020)
- [Invariant Risk Minimization](https://mlsanity.com/paper/1907.02893/) (2019)
- [WILDS: A Benchmark of in-the-Wild Distribution Shifts](https://mlsanity.com/paper/2012.07421/) (2020)
- [In Search of Lost Domain Generalization](https://mlsanity.com/paper/2007.01434/) (2020)
- [Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss](https://mlsanity.com/paper/1906.07413/) (2019)
- [Exact solutions to the nonlinear dynamics of learning in deep linear neural networks](https://mlsanity.com/paper/1312.6120/) (2013)
- [Out-of-Distribution Generalization via Risk Extrapolation (REx)](https://mlsanity.com/paper/2003.00688/) (2020)
- [SphereFace: Deep Hypersphere Embedding for Face Recognition](https://mlsanity.com/paper/1704.08063/) (2017)
- [The Implicit Bias of Gradient Descent on Separable Data](https://mlsanity.com/paper/1710.10345/) (2017)
- [Just Train Twice: Improving Group Robustness without Training Group Information](https://mlsanity.com/paper/2107.09044/) (2021)
- [Nuanced Metrics for Measuring Unintended Bias with Real Data for Text Classification](https://mlsanity.com/paper/1903.04561/) (2019)
- [The Pitfalls of Simplicity Bias in Neural Networks](https://mlsanity.com/paper/2006.07710/) (2020)
- [An Investigation of Why Overparameterization Exacerbates Spurious Correlations](https://mlsanity.com/paper/2005.04345/) (2020)
- [Synthetic and Natural Noise Both Break Neural Machine Translation](https://mlsanity.com/paper/1711.02173/) (2017)
- [Environment Inference for Invariant Learning](https://mlsanity.com/paper/2010.07249/) (2020)
- [Gradient Starvation: A Learning Proclivity in Neural Networks](https://mlsanity.com/paper/2011.09468/) (2020)
- [High-dimensional dynamics of generalization error in neural networks](https://mlsanity.com/paper/1710.03667/) (2017)
- [Invariance Principle Meets Information Bottleneck for Out-of-Distribution Generalization](https://mlsanity.com/paper/2106.06607/) (2021)
- [No Subclass Left Behind: Fine-Grained Robustness in Coarse-Grained Classification Problems](https://mlsanity.com/paper/2011.12945/) (2020)
- [Deep learning generalizes because the parameter-function map is biased towards simple functions](https://mlsanity.com/paper/1805.08522/) (2018)
- [Simple data balancing achieves competitive worst-group-accuracy](https://mlsanity.com/paper/2110.14503/) (2021)
- [Correct-N-Contrast: A Contrastive Approach for Improving Robustness to Spurious Correlations](https://mlsanity.com/paper/2203.01517/) (2022)
- [Towards Explaining the Regularization Effect of Initial Large Learning Rate in Training Neural Networks](https://mlsanity.com/paper/1907.04595/) (2019)
- [Understanding the Failure Modes of Out-of-Distribution Generalization](https://mlsanity.com/paper/2010.15775/) (2020)
- [The Convergence Rate of Neural Networks for Learned Functions of Different Frequencies](https://mlsanity.com/paper/1906.00425/) (2019)
- [On Calibration and Out-of-domain Generalization](https://mlsanity.com/paper/2102.10395/) (2021)
- [What shapes feature representations? Exploring datasets, architectures, and training](https://mlsanity.com/paper/2006.12433/) (2020)
- [Implicit Regularization of Discrete Gradient Dynamics in Linear Neural Networks](https://mlsanity.com/paper/1904.13262/) (2019)

## Built on by (most cited first, 30 of 30)

- [On the Foundations of Shortcut Learning](https://mlsanity.com/paper/2310.16228/) (2023)
- [On the Challenges and Opportunities in Generative AI](https://mlsanity.com/paper/2403.00025/) (2024)
- [The Pitfalls of Memorization: When Memorization Hurts Generalization](https://mlsanity.com/paper/2412.07684/) (2024)
- [Shortcut Learning in Medical Image Segmentation](https://mlsanity.com/paper/2403.06748/) (2024)
- [Data Augmentations for Improved (Large) Language Model Generalization](https://mlsanity.com/paper/2310.12803/) (2023)
- [Do We Always Need the Simplicity Bias? Looking for Optimal Inductive Biases in the Wild](https://mlsanity.com/paper/2503.10065/) (2025)
- [To Use or not to Use Muon: How Simplicity Bias in Optimizers Matters](https://mlsanity.com/paper/2603.00742/) (2026)
- [Tackling Data Heterogeneity in Federated Learning via Loss Decomposition](https://mlsanity.com/paper/2408.12300/) (2024)
- [Enhancing Domain Adaptation through Prompt Gradient Alignment](https://mlsanity.com/paper/2406.09353/) (2024)
- [Generative Classifiers Avoid Shortcut Solutions](https://mlsanity.com/paper/2512.25034/) (2025)
- [Towards Formalizing Spuriousness of Biased Datasets Using Partial Information Decomposition](https://mlsanity.com/paper/2407.00482/) (2024)
- [CHARM3R: Towards Unseen Camera Height Robust Monocular 3D Detector](https://mlsanity.com/paper/2508.11185/) (2025)
- [Data Debiasing with Datamodels (D3M): Improving Subgroup Robustness via Data Selection](https://mlsanity.com/paper/2406.16846/) (2024)
- [Changing the Training Data Distribution to Reduce Simplicity Bias Improves In-distribution Generalization](https://mlsanity.com/paper/2404.17768/) (2024)
- [Explanations that reveal all through the definition of encoding](https://mlsanity.com/paper/2411.02664/) (2024)
- [Mitigating Shortcut Learning with InterpoLated Learning](https://mlsanity.com/paper/2507.05527/) (2025)
- [Large Learning Rates Simultaneously Achieve Robustness to Spurious Correlations and Compressibility](https://mlsanity.com/paper/2507.17748/) (2025)
- [Do Students Debias Like Teachers? On the Distillability of Bias Mitigation Methods](https://mlsanity.com/paper/2510.26038/) (2025)
- [Can Transformers Really Do It All? On the Compatibility of Inductive Biases Across Tasks](https://mlsanity.com/paper/2607.17624/) (2026)
- [FairDropout: Using Example-Tied Dropout to Enhance Generalization of Minority Groups](https://mlsanity.com/paper/2502.06695/) (2025)
- [MiMu: Mitigating Multiple Shortcut Learning Behavior of Transformers](https://mlsanity.com/paper/2504.10551/) (2025)
- [Fine-Grained Class-Conditional Distribution Balancing for Debiased Learning](https://mlsanity.com/paper/2505.06831/) (2025)
- [On Measuring Localization of Shortcuts in Deep Networks](https://mlsanity.com/paper/2510.26560/) (2025)
- [Unsupervised Multi-modal Feature Alignment for Time Series Representation Learning](https://mlsanity.com/paper/2312.05698/) (2023)
- [ALVIN: Active Learning Via INterpolation](https://mlsanity.com/paper/2410.08972/) (2024)
- [Open-Set Domain Adaptation Under Background Distribution Shift: Challenges and A Provably Efficient Solution](https://mlsanity.com/paper/2512.01152/) (2025)
- [Shortcut Features as Top Eigenfunctions of NTK: A Linear Neural Network Case and More](https://mlsanity.com/paper/2602.03066/) (2026)
- [Revisiting Autoregressive Models for Generative Image Classification](https://mlsanity.com/paper/2603.19122/) (2026)
- [FeDMRA: Federated Incremental Learning with Dynamic Memory Replay Allocation](https://mlsanity.com/paper/2603.28455/) (2026)
- [Whitening Improves Robustness to Spurious Correlations in Linear Probes](https://mlsanity.com/paper/2609.39177/) (2026)

## Abstract (arXiv)

Common explanations for shortcut learning assume that the shortcut improves prediction under the training distribution but not in the test distribution. Thus, models trained via the typical gradient-based optimization of cross-entropy, which we call default-ERM, utilize the shortcut. However, even when the stable feature determines the label in the training distribution and the shortcut does not provide any additional information, like in perception tasks, default-ERM still exhibits shortcut learning. Why are such solutions preferred when the loss for default-ERM can be driven to zero using the stable feature alone? By studying a linear perception task, we show that default-ERM's preference for maximizing the margin leads to models that depend more on the shortcut than the stable feature, even without overparameterization. This insight suggests that default-ERM's implicit inductive bias towards max-margin is unsuitable for perception tasks. Instead, we develop an inductive bias toward uniform margins and show that this bias guarantees dependence only on the perfect stable feature in the linear perception task. We develop loss functions that encourage uniform-margin solutions, called margin control (MARG-CTRL). MARG-CTRL mitigates shortcut learning on a variety of vision and language tasks, showing that better inductive biases can remove the need for expensive two-stage shortcut-mitigating methods in perception tasks.