# Can the state of relevant neurons in a deep neural networks serve as indicators for detecting adversarial attacks?

Source: [Can the state of relevant neurons in a deep neural networks serve as indicators for detecting adversarial attacks?](<https://www.sota2.com/research/paper/68fa68a91e209131cea1d927>)

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

Data updated: 2025-10-25T22:42:59.203Z

Authors: Roger Granda, Tinne Tuytelaars, Jose Oramas

Published: 2020-10-29T22:31:42.000Z

## Abstract

We present a method for adversarial attack detection based on the inspection of a sparse set of neurons. We follow the hypothesis that adversarial attacks introduce imperceptible perturbations in the input and that these perturbations change the state of neurons relevant for the concepts modelled by the attacked model. Therefore, monitoring the status of these neurons would enable the detection of adversarial attacks. Focusing on the image classification task, our method identifies neurons that are relevant for the classes predicted by the model. A deeper qualitative inspection of these sparse set of neurons indicates that their state changes in the presence of adversarial samples. Moreover, quantitative results from our empirical evaluation indicate that our method is capable of recognizing adversarial samples, produced by state-of-the-art attack methods, with comparable accuracy to that of state-of-the-art detectors.

## 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/2010.15974>)

[PDF](<https://arxiv.org/pdf/2010.15974.pdf>)