FAME: Formal Abstract Minimal Explanation for Neural Networks
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918383363031040 |
|---|---|
| author | Boumazouza, Ryma Elsaleh, Raya Ducoffe, Melanie Bassan, Shahaf Katz, Guy |
| author_facet | Boumazouza, Ryma Elsaleh, Raya Ducoffe, Melanie Bassan, Shahaf Katz, Guy |
| contents | We propose FAME (Formal Abstract Minimal Explanations), a new class of abductive explanations grounded in abstract interpretation. FAME is the first method to scale to large neural networks while reducing explanation size. Our main contribution is the design of dedicated perturbation domains that eliminate the need for traversal order. FAME progressively shrinks these domains and leverages LiRPA-based bounds to discard irrelevant features, ultimately converging to a formal abstract minimal explanation. To assess explanation quality, we introduce a procedure that measures the worst-case distance between an abstract minimal explanation and a true minimal explanation. This procedure combines adversarial attacks with an optional VERIX+ refinement step. We benchmark FAME against VERIX+ and demonstrate consistent gains in both explanation size and runtime on medium- to large-scale neural networks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_10661 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | FAME: Formal Abstract Minimal Explanation for Neural Networks Boumazouza, Ryma Elsaleh, Raya Ducoffe, Melanie Bassan, Shahaf Katz, Guy Artificial Intelligence Machine Learning We propose FAME (Formal Abstract Minimal Explanations), a new class of abductive explanations grounded in abstract interpretation. FAME is the first method to scale to large neural networks while reducing explanation size. Our main contribution is the design of dedicated perturbation domains that eliminate the need for traversal order. FAME progressively shrinks these domains and leverages LiRPA-based bounds to discard irrelevant features, ultimately converging to a formal abstract minimal explanation. To assess explanation quality, we introduce a procedure that measures the worst-case distance between an abstract minimal explanation and a true minimal explanation. This procedure combines adversarial attacks with an optional VERIX+ refinement step. We benchmark FAME against VERIX+ and demonstrate consistent gains in both explanation size and runtime on medium- to large-scale neural networks. |
| title | FAME: Formal Abstract Minimal Explanation for Neural Networks |
| topic | Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2603.10661 |