An empirical study of fault localisation techniques for deep neural networks
Additional information
Authors
Type
Journal Article
Year
2025
Language
English
Abstract
With the increased popularity of Deep Neural Networks (DNNs), increases also the need for tools to assist developers in the DNN implementation, testing and debugging process. Several approaches have been proposed that automatically analyse and localise potential faults in DNNs under test. In this work, we evaluate and compare existing state-of-the-art fault localisation techniques, which operate based on both dynamic and static analysis of the DNN. The evaluation is performed on a benchmark consisting of both real faults obtained from bug reporting platforms and faulty models produced by a mutation tool. Our findings indicate that the usage of a single, specific ground truth (e.g. the human-defined one) for the evaluation of DNN fault localisation tools results in pretty low performance (maximum average recall of 0.33 and precision of 0.21). However, such figures increase when considering alternative, equivalent patches that exist for a given faulty DNN. The results indicate that DeepFD is the most effective tool, achieving an average recall of 0.55 and a precision of 0.37 on our benchmark.
Keywords
Deep learning, Real faults, Fault localisation
Journal
Empirical Software Engineering
Volume
30
Number ( Month )
5
Pages (or article number)
124
ISSN
1382-3256, 1573-7616
Diffusion
License
CC BY
Visibility
Public
Status open access
Hybrid
