English

SIDE: Sparse Information Disentanglement for Explainable Artificial Intelligence

Computer Vision and Pattern Recognition 2025-07-28 v1 Artificial Intelligence Machine Learning

Abstract

Understanding the decisions made by deep neural networks is essential in high-stakes domains such as medical imaging and autonomous driving. Yet, these models often lack transparency, particularly in computer vision. Prototypical-parts-based neural networks have emerged as a promising solution by offering concept-level explanations. However, most are limited to fine-grained classification tasks, with few exceptions such as InfoDisent. InfoDisent extends prototypical models to large-scale datasets like ImageNet, but produces complex explanations. We introduce Sparse Information Disentanglement for Explainability (SIDE), a novel method that improves the interpretability of prototypical parts through a dedicated training and pruning scheme that enforces sparsity. Combined with sigmoid activations in place of softmax, this approach allows SIDE to associate each class with only a small set of relevant prototypes. Extensive experiments show that SIDE matches the accuracy of existing methods while reducing explanation size by over 90%90\%, substantially enhancing the understandability of prototype-based explanations.

Keywords

Cite

@article{arxiv.2507.19321,
  title  = {SIDE: Sparse Information Disentanglement for Explainable Artificial Intelligence},
  author = {Viktar Dubovik and Łukasz Struski and Jacek Tabor and Dawid Rymarczyk},
  journal= {arXiv preprint arXiv:2507.19321},
  year   = {2025}
}
R2 v1 2026-07-01T04:18:57.212Z