SpArX: sparse argumentative explanations for neural networks

Ayoobi, Hamed, Potyka, Nico and Toni, Francesca 2023. SpArX: sparse argumentative explanations for neural networks. Presented at: 26th European Conference on Artificial Intelligence (ECAI), 30 September - 04 October 2023. ECAI 2023. Frontiers in Artificial Intelligence and Applications , vol.372 IOS Press, pp. 149-156. 10.3233/FAIA230265

Preview

PDF - Published Version Available under License Creative Commons Attribution Non-commercial. Download (753kB) | Preview

Abstract

Neural networks (NNs) have various applications in AI, but explaining their decisions remains challenging. Existing approaches often focus on explaining how changing individual inputs affects NNs’ outputs. However, an explanation that is consistent with the input-output behaviour of an NN is not necessarily faithful to the actual mechanics thereof. In this paper, we exploit relationships between multi-layer perceptrons (MLPs) and quantitative argumentation frameworks (QAFs) to create argumentative explanations for the mechanics of MLPs. Our SpArX method first sparsifies the MLP while maintaining as much of the original structure as possible. It then translates the sparse MLP into an equivalent QAF to shed light on the underlying decision process of the MLP, producing global and/or local explanations. We demonstrate experimentally that SpArX can give more faithful explanations than existing approaches, while simultaneously providing deeper insights into the actual reasoning process of MLPs.

Item Type:	Conference or Workshop Item (Paper)
Status:	Published
Schools:	Computer Science & Informatics
Publisher:	IOS Press
Date of First Compliant Deposit:	19 October 2023
Last Modified:	08 Nov 2023 10:00
URI:	https://orca.cardiff.ac.uk/id/eprint/163315

Actions (repository staff only)

Edit Item