# Understanding Classifier Mistakes with Generative Models

arXiv: https://arxiv.org/abs/2010.02364 · HTML: https://mlsanity.com/paper/2010.02364/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 5 Oct 2020.

## Built on by (most cited first, 4 of 4)

- [Measuring Unintended Memorisation of Unique Private Features in Neural Networks](https://mlsanity.com/paper/2202.08099/) (2022)
- [Unintended memorisation of unique features in neural networks](https://mlsanity.com/paper/2205.10079/) (2022)
- [Towards Reliable Time Series Forecasting under Future Uncertainty: Ambiguity and Novelty Rejection Mechanisms](https://mlsanity.com/paper/2503.19656/) (2025)
- [Constraining Representations Yields Models That Know What They Don't Know](https://mlsanity.com/paper/2208.14488/) (2022)

## Abstract (arXiv)

Although deep neural networks are effective on supervised learning tasks, they have been shown to be brittle. They are prone to overfitting on their training distribution and are easily fooled by small adversarial perturbations. In this paper, we leverage generative models to identify and characterize instances where classifiers fail to generalize. We propose a generative model of the features extracted by a classifier, and show using rigorous hypothesis testing that errors tend to occur when features are assigned low-probability by our model. From this observation, we develop a detection criteria for samples on which a classifier is likely to fail at test time. In particular, we test against three different sources of classification failures: mistakes made on the test set due to poor model generalization, adversarial samples and out-of-distribution samples. Our approach is agnostic to class labels from the training set which makes it applicable to models trained in a semi-supervised way.