# Can Biases in ImageNet Models Explain Generalization?

arXiv: https://arxiv.org/abs/2404.01509 · HTML: https://mlsanity.com/paper/2404.01509/

## Key facts

- Cited by 20 papers in the map (data through 14 Oct 2026).
- Builds on 52 of its 69 references that are in the map.
- First posted on arXiv 1 Apr 2024.

## Builds on (most cited first, 30 of 52)

- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [ImageNet Large Scale Visual Recognition Challenge](https://mlsanity.com/paper/1409.0575/) (2014)
- [Masked Autoencoders Are Scalable Vision Learners](https://mlsanity.com/paper/2111.06377/) (2021)
- [Emerging Properties in Self-Supervised Vision Transformers](https://mlsanity.com/paper/2104.14294/) (2021)
- [Towards Deep Learning Models Resistant to Adversarial Attacks](https://mlsanity.com/paper/1706.06083/) (2017)
- [Intriguing properties of neural networks](https://mlsanity.com/paper/1312.6199/) (2013)
- [Training data-efficient image transformers & distillation through attention](https://mlsanity.com/paper/2012.12877/) (2020)
- [Benchmarking Neural Network Robustness to Common Corruptions and Perturbations](https://mlsanity.com/paper/1903.12261/) (2019)
- [Unsupervised Learning of Visual Features by Contrasting Cluster Assignments](https://mlsanity.com/paper/2006.09882/) (2020)
- [The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalization](https://mlsanity.com/paper/2006.16241/) (2020)
- [An Empirical Study of Training Self-Supervised Vision Transformers](https://mlsanity.com/paper/2104.02057/) (2021)
- [ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness](https://mlsanity.com/paper/1811.12231/) (2018)
- [Natural Adversarial Examples](https://mlsanity.com/paper/1907.07174/) (2019)
- [Big Self-Supervised Models are Strong Semi-Supervised Learners](https://mlsanity.com/paper/2006.10029/) (2020)
- [Learning Robust Global Representations by Penalizing Local Predictive Power](https://mlsanity.com/paper/1905.13549/) (2019)
- [Adversarial Examples Are Not Bugs, They Are Features](https://mlsanity.com/paper/1905.02175/) (2019)
- [AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty](https://mlsanity.com/paper/1912.02781/) (2019)
- [Robustness May Be at Odds with Accuracy](https://mlsanity.com/paper/1805.12152/) (2018)
- [Scaling Vision Transformers to 22 Billion Parameters](https://mlsanity.com/paper/2302.05442/) (2023)
- [Evasion Attacks against Machine Learning at Test Time](https://mlsanity.com/paper/1708.06131/) (2017)
- [How to train your ViT? Data, Augmentation, and Regularization in Vision Transformers](https://mlsanity.com/paper/2106.10270/) (2021)
- [Measuring Robustness to Natural Distribution Shifts in Image Classification](https://mlsanity.com/paper/2007.00644/) (2020)
- [Intriguing Properties of Vision Transformers](https://mlsanity.com/paper/2105.10497/) (2021)
- [ResNet strikes back: An improved training procedure in timm](https://mlsanity.com/paper/2110.00476/) (2021)
- [Do Adversarially Robust ImageNet Models Transfer Better?](https://mlsanity.com/paper/2007.08489/) (2020)
- [High Frequency Component Helps Explain the Generalization of Convolutional Neural Networks](https://mlsanity.com/paper/1905.13545/) (2019)
- [A Fourier Perspective on Model Robustness in Computer Vision](https://mlsanity.com/paper/1906.08988/) (2019)
- [When Vision Transformers Outperform ResNets without Pre-training or Strong Data Augmentations](https://mlsanity.com/paper/2106.01548/) (2021)
- [Billion-scale semi-supervised learning for image classification](https://mlsanity.com/paper/1905.00546/) (2019)

## Built on by (most cited first, 20 of 20)

- [Wavelet Convolutions for Large Receptive Fields](https://mlsanity.com/paper/2407.05848/) (2024)
- [TinyViM: Frequency Decoupling for Tiny Hybrid Vision Mamba](https://mlsanity.com/paper/2411.17473/) (2024)
- [FreqDebias: Towards Generalizable Deepfake Detection via Consistency-Driven Frequency Debiasing](https://mlsanity.com/paper/2509.22412/) (2025)
- [ImageNet-trained CNNs are not biased towards texture: Revisiting feature reliance through controlled suppression](https://mlsanity.com/paper/2509.20234/) (2025)
- [When Pretty Isn't Useful: Investigating Why Modern Text-to-Image Models Fail as Reliable Training Data Generators](https://mlsanity.com/paper/2602.19946/) (2026)
- [Do ImageNet-trained models learn shortcuts? The impact of frequency shortcuts on generalization](https://mlsanity.com/paper/2503.03519/) (2025)
- [3D Wavelet Convolutions with Extended Receptive Fields for Hyperspectral Image Classification](https://mlsanity.com/paper/2504.10795/) (2025)
- [SciPostGen: Bridging the Gap between Scientific Papers and Poster Layouts](https://mlsanity.com/paper/2511.22490/) (2025)
- [How Do Training Methods Influence the Utilization of Vision Models?](https://mlsanity.com/paper/2410.14470/) (2024)
- [Attention IoU: Examining Biases in CelebA using Attention Maps](https://mlsanity.com/paper/2503.19846/) (2025)
- [Automated Detection of Visual Attribute Reliance with a Self-Reflective Agent](https://mlsanity.com/paper/2510.21704/) (2025)
- [Low-Biased General Annotated Dataset Generation](https://mlsanity.com/paper/2412.10831/) (2024)
- [One Model, Many Behaviors: Training-Induced Effects on Out-of-Distribution Detection](https://mlsanity.com/paper/2601.10836/) (2026)
- [On the Reliability of Cue Conflict and Beyond](https://mlsanity.com/paper/2603.10834/) (2026)
- [Vision Transformer Neural Architecture Search for Out-of-Distribution Generalization: Benchmark and Insights](https://mlsanity.com/paper/2501.03782/) (2025)
- [M3Depth: Wavelet-Enhanced Depth Estimation on Mars via Mutual Boosting of Dual-Modal Data](https://mlsanity.com/paper/2505.14159/) (2025)
- [ShapeY: A Principled Framework for Measuring Shape Recognition Capacity via Nearest-Neighbor Matching](https://mlsanity.com/paper/2604.25065/) (2026)
- [Low-Frequency Shortcuts in Texture-Driven Visual Learning](https://mlsanity.com/paper/2606.03493/) (2026)
- [Robustness Emerges Early in Training Dynamics, but Is Not Preserved](https://mlsanity.com/paper/2608.04442/) (2026)
- [Suppress and Diversify: Refining Robust Pathways for Corruption Robustness](https://mlsanity.com/paper/2608.06712/) (2026)

## Abstract (arXiv)

The robust generalization of models to rare, in-distribution (ID) samples drawn from the long tail of the training distribution and to out-of-training-distribution (OOD) samples is one of the major challenges of current deep learning methods. For image classification, this manifests in the existence of adversarial attacks, the performance drops on distorted images, and a lack of generalization to concepts such as sketches. The current understanding of generalization in neural networks is very limited, but some biases that differentiate models from human vision have been identified and might be causing these limitations. Consequently, several attempts with varying success have been made to reduce these biases during training to improve generalization. We take a step back and sanity-check these attempts. Fixing the architecture to the well-established ResNet-50, we perform a large-scale study on 48 ImageNet models obtained via different training methods to understand how and if these biases - including shape bias, spectral biases, and critical bands - interact with generalization. Our extensive study results reveal that contrary to previous findings, these biases are insufficient to accurately predict the generalization of a model holistically. We provide access to all checkpoints and evaluation code at https://github.com/paulgavrikov/biases_vs_generalization