XMutant: XAI-based Fuzzing for Deep Learning Systems

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chen, Xingcheng, Biagiola, Matteo, Riccio, Vincenzo, d'Amorim, Marcelo, Stocco, Andrea
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908677962727424
author Chen, Xingcheng
Biagiola, Matteo
Riccio, Vincenzo
d'Amorim, Marcelo
Stocco, Andrea
author_facet Chen, Xingcheng
Biagiola, Matteo
Riccio, Vincenzo
d'Amorim, Marcelo
Stocco, Andrea
contents Semantic-based test generators are widely used to produce failure-inducing inputs for Deep Learning (DL) systems. They typically generate challenging test inputs by applying random perturbations to input semantic concepts until a failure is found or a timeout is reached. However, such randomness may hinder them from efficiently achieving their goal. This paper proposes XMutant, a technique that leverages explainable artificial intelligence (XAI) techniques to generate challenging test inputs. XMutant uses the local explanation of the input to inform the fuzz testing process and effectively guide it toward failures of the DL system under test. We evaluated different configurations of XMutant in triggering failures for different DL systems both for model-level (sentiment analysis, digit recognition) and system-level testing (advanced driving assistance). Our studies showed that XMutant enables more effective and efficient test generation by focusing on the most impactful parts of the input. XMutant generates up to 125% more failure-inducing inputs compared to an existing baseline, up to 7X faster. We also assessed the validity of these inputs, maintaining a validation rate above 89%, according to automated and human validators.
format Preprint
id arxiv_https___arxiv_org_abs_2503_07222
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle XMutant: XAI-based Fuzzing for Deep Learning Systems
Chen, Xingcheng
Biagiola, Matteo
Riccio, Vincenzo
d'Amorim, Marcelo
Stocco, Andrea
Software Engineering
Semantic-based test generators are widely used to produce failure-inducing inputs for Deep Learning (DL) systems. They typically generate challenging test inputs by applying random perturbations to input semantic concepts until a failure is found or a timeout is reached. However, such randomness may hinder them from efficiently achieving their goal. This paper proposes XMutant, a technique that leverages explainable artificial intelligence (XAI) techniques to generate challenging test inputs. XMutant uses the local explanation of the input to inform the fuzz testing process and effectively guide it toward failures of the DL system under test. We evaluated different configurations of XMutant in triggering failures for different DL systems both for model-level (sentiment analysis, digit recognition) and system-level testing (advanced driving assistance). Our studies showed that XMutant enables more effective and efficient test generation by focusing on the most impactful parts of the input. XMutant generates up to 125% more failure-inducing inputs compared to an existing baseline, up to 7X faster. We also assessed the validity of these inputs, maintaining a validation rate above 89%, according to automated and human validators.
title XMutant: XAI-based Fuzzing for Deep Learning Systems
topic Software Engineering
url https://arxiv.org/abs/2503.07222