**arXiv ID:** 2403.09863 **Authors:** Maciej Satkiewicz **Published:** 2024-03-14T20:50:03Z **Abstract:** Deep neural networks learn fragile "shortcut" features, rendering them difficult to interpret (black box) and vulnerable to adversarial attacks. This paper proposes semantic features as a general architectural solution to this problem. The main idea is to make features locality-sensitive in the adequate semantic topology of the domain, thus introducing a strong regularization. The proof o...
Installs into .claude/skills of the current project.
Sponsors
SSkills DirectorySkills Directory
Know which skills are safe — weekly.
SSkills DirectorySkills Directory
Is that skill safe to install?
Are you the author of Towards White Box Deep Learning?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/hiyenwong-towards-white-box-deep-learning)
# Towards White Box Deep Learning
**arXiv ID:** 2403.09863
**Authors:** Maciej Satkiewicz
**Published:** 2024-03-14T20:50:03Z
**Abstract:**
Deep neural networks learn fragile "shortcut" features, rendering them difficult to interpret (black box) and vulnerable to adversarial attacks. This paper proposes semantic features as a general architectural solution to this problem. The main idea is to make features locality-sensitive in the adequate semantic topology of the domain, thus introducing a strong regularization. The proof of concept network is lightweight, inherently interpretable and achieves almost human-level adversarial test metrics - with no adversarial training! These results and the general nature of the approach warrant further research on semantic features. The code is available at https://github.com/314-Foundation/white-box-nn
## Skill Description
This skill is generated from the arXiv paper: Towards White Box Deep Learning (2403.09863).
## How to Use
[To be filled in by the user or by future automation]
## References
- [arXiv:2403.09863](http://arxiv.org/abs/2403.09863v5)