SIDE: Sparse Information Disentanglement for Explainable Artificial Intelligence

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Dubovik, Viktar, Struski, Łukasz, Tabor, Jacek, Rymarczyk, Dawid
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918104344297472
author Dubovik, Viktar
Struski, Łukasz
Tabor, Jacek
Rymarczyk, Dawid
author_facet Dubovik, Viktar
Struski, Łukasz
Tabor, Jacek
Rymarczyk, Dawid
contents 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 $90\%$, substantially enhancing the understandability of prototype-based explanations.
format Preprint
id arxiv_https___arxiv_org_abs_2507_19321
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SIDE: Sparse Information Disentanglement for Explainable Artificial Intelligence
Dubovik, Viktar
Struski, Łukasz
Tabor, Jacek
Rymarczyk, Dawid
Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
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 $90\%$, substantially enhancing the understandability of prototype-based explanations.
title SIDE: Sparse Information Disentanglement for Explainable Artificial Intelligence
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2507.19321