Enhancing Vision Transformer Explainability Using Artificial Astrocytes
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909625383649280 |
|---|---|
| author | Echevarrieta-Catalan, Nicolas Ribas-Rodriguez, Ana Cedron, Francisco Schwartz, Odelia Aguiar-Pulido, Vanessa |
| author_facet | Echevarrieta-Catalan, Nicolas Ribas-Rodriguez, Ana Cedron, Francisco Schwartz, Odelia Aguiar-Pulido, Vanessa |
| contents | Machine learning models achieve high precision, but their decision-making processes often lack explainability. Furthermore, as model complexity increases, explainability typically decreases. Existing efforts to improve explainability primarily involve developing new eXplainable artificial intelligence (XAI) techniques or incorporating explainability constraints during training. While these approaches yield specific improvements, their applicability remains limited. In this work, we propose the Vision Transformer with artificial Astrocytes (ViTA). This training-free approach is inspired by neuroscience and enhances the reasoning of a pretrained deep neural network to generate more human-aligned explanations. We evaluated our approach employing two well-known XAI techniques, Grad-CAM and Grad-CAM++, and compared it to a standard Vision Transformer (ViT). Using the ClickMe dataset, we quantified the similarity between the heatmaps produced by the XAI techniques and a (human-aligned) ground truth. Our results consistently demonstrate that incorporating artificial astrocytes enhances the alignment of model explanations with human perception, leading to statistically significant improvements across all XAI techniques and metrics utilized. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_21513 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Enhancing Vision Transformer Explainability Using Artificial Astrocytes Echevarrieta-Catalan, Nicolas Ribas-Rodriguez, Ana Cedron, Francisco Schwartz, Odelia Aguiar-Pulido, Vanessa Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning Machine learning models achieve high precision, but their decision-making processes often lack explainability. Furthermore, as model complexity increases, explainability typically decreases. Existing efforts to improve explainability primarily involve developing new eXplainable artificial intelligence (XAI) techniques or incorporating explainability constraints during training. While these approaches yield specific improvements, their applicability remains limited. In this work, we propose the Vision Transformer with artificial Astrocytes (ViTA). This training-free approach is inspired by neuroscience and enhances the reasoning of a pretrained deep neural network to generate more human-aligned explanations. We evaluated our approach employing two well-known XAI techniques, Grad-CAM and Grad-CAM++, and compared it to a standard Vision Transformer (ViT). Using the ClickMe dataset, we quantified the similarity between the heatmaps produced by the XAI techniques and a (human-aligned) ground truth. Our results consistently demonstrate that incorporating artificial astrocytes enhances the alignment of model explanations with human perception, leading to statistically significant improvements across all XAI techniques and metrics utilized. |
| title | Enhancing Vision Transformer Explainability Using Artificial Astrocytes |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2505.21513 |