# Detecting Overfitting via Adversarial Examples

Source: [Detecting Overfitting via Adversarial Examples](<https://www.sota2.com/research/paper/68fa66af1e209131cea11c1d>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-28T13:56:41.021Z

Authors: Roman Werpachowski, Andr\\'as Gy\\"orgy, Csaba Szepesv\\'ari

Published: 2019-03-06T13:49:18.000Z

## Abstract

The repeated community-wide reuse of test sets in popular benchmark problems raises doubts about the credibility of reported test-error rates. Verifying whether a learned model is overfitted to a test set is challenging as independent test sets drawn from the same data distribution are usually unavailable, while other test sets may introduce a distribution shift. We propose a new hypothesis test that uses only the original test data to detect overfitting. It utilizes a new unbiased error estimate that is based on adversarial examples generated from the test data and importance weighting. Overfitting is detected if this error estimate is sufficiently different from the original test error rate. We develop a specialized variant of our test for multiclass image classification, and apply it to testing overfitting of recent models to the popular ImageNet benchmark. Our method correctly indicates overfitting of the trained model to the training set, but is not able to detect any overfitting to the test set, in line with other recent work on this topic.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/1903.02380>)

[PDF](<https://arxiv.org/pdf/1903.02380.pdf>)
