FAME: Formal Abstract Minimal Explanation for Neural Networks

• URL: http://arxiv.org/abs/2603.10661v1
• Date: Wed, 11 Mar 2026 11:19:21 GMT
• Title: FAME: Formal Abstract Minimal Explanation for Neural Networks
• Abstract summary: We propose FAME, a new class of abductive explanations grounded in abstract interpretation.<n>FAME is the first method to scale to large neural networks while reducing explanation size.<n>We benchmark FAME against VERIX+ and demonstrate consistent gains in both explanation size and runtime.
• Score: 5.089700375729287
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: 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.

Related papers

• Efficient Reasoning via Thought Compression for Language Segmentation [28.138192925569317]
  Chain-of-thought (CoT) reasoning has significantly improved the performance of large multimodal models in language-guided segmentation.<n>We introduce WISE, a novel paradigm for efficient reasoning guided by the principle of textitthinking twice -- once for learning, once for speed.<n>We show WISE achieves state-of-the-art zero-shot performance on the ReasonSeg benchmark with 58.3 cIoU, while reducing the average reasoning length by nearly textbf5$times$ -- from 112 to just 23 tokens.
  arXiv  Detail & Related papers  (2026-04-02T13:45:01Z)
• Slice and Explain: Logic-Based Explanations for Neural Networks through Domain Slicing [3.847524185858632]
  This paper proposes an approach leveraging domain slicing to facilitate explanation generation for NNs.<n>By reducing the complexity of logical constraints through slicing, we decrease explanation time by up to 40% less time.
  arXiv  Detail & Related papers  (2026-02-25T17:01:52Z)
• SIDE: Sparse Information Disentanglement for Explainable Artificial Intelligence [9.975642488603937]
  Prototypical-parts-based neural networks have emerged as a promising solution by offering concept-level explanations.<n>We introduce Sparse Information Disentanglement for Explainability (SIDE), a novel method that improves the interpretability of prototypical parts.
  arXiv  Detail & Related papers  (2025-07-25T14:34:15Z)
• A Lightweight Deep Exclusion Unfolding Network for Single Image Reflection Removal [68.0573194557999]
  Single Image Reflection Removal (SIRR) is a canonical blind source separation problem.<n>We propose a novel Deep Exclusion unfolding Network (DExNet) for SIRR.<n>DExNet is constructed by unfolding and parameterizing a simple iterative Sparse and Auxiliary Feature Update (i-SAFU) algorithm.
  arXiv  Detail & Related papers  (2025-03-03T07:54:27Z)
• Efficient Contrastive Explanations on Demand [7.109897082275965]
  This paper proposes novel algorithms to compute the so-called contrastive explanations for machine learning models.<n>The paper also proposes novel algorithms for listing explanations and finding smallest contrastive explanations.
  arXiv  Detail & Related papers  (2024-12-24T08:24:10Z)
• Semantic Loss Functions for Neuro-Symbolic Structured Prediction [74.18322585177832]
  We discuss the semantic loss, which injects knowledge about such structure, defined symbolically, into training. It is agnostic to the arrangement of the symbols, and depends only on the semantics expressed thereby. It can be combined with both discriminative and generative neural models.
  arXiv  Detail & Related papers  (2024-05-12T22:18:25Z)
• What Sketch Explainability Really Means for Downstream Tasks [121.85845328822641]
  We explore the unique modality of sketch for explainability, emphasising the profound impact of human strokes. We propose a lightweight and portable explainability solution -- a seamless plugin that integrates effortlessly with any pre-trained model.
  arXiv  Detail & Related papers  (2024-03-14T15:22:33Z)
• Interpretations Steered Network Pruning via Amortized Inferred Saliency Maps [85.49020931411825]
  Convolutional Neural Networks (CNNs) compression is crucial to deploying these models in edge devices with limited resources. We propose to address the channel pruning problem from a novel perspective by leveraging the interpretations of a model to steer the pruning process. We tackle this challenge by introducing a selector model that predicts real-time smooth saliency masks for pruned models.
  arXiv  Detail & Related papers  (2022-09-07T01:12:11Z)
• Connection-minimal Abduction in EL via Translation to FOL -- Technical Report [12.90382979353427]
  We show how to compute a class of connection-minimal hypotheses in a sound and complete way. Our technique is based on a translation to first-order logic, and constructs hypotheses based on prime implicates.
  arXiv  Detail & Related papers  (2022-05-17T15:50:27Z)
• How to Explain Neural Networks: A perspective of data space division [2.4499092754102874]
  Interpretability of algorithms represented by deep learning has been yet an open problem. We discuss the shortcomings of the existing explainable method based on the two attributes of explanation, which are called completeness and explicitness. Based on the perspective of the data space division, the principle of complete local interpretable model-agnostic explanations (CLIMEP) is proposed in this paper.
  arXiv  Detail & Related papers  (2021-05-17T13:43:37Z)
• Effective Version Space Reduction for Convolutional Neural Networks [61.84773892603885]
  In active learning, sampling bias could pose a serious inconsistency problem and hinder the algorithm from finding the optimal hypothesis. We examine active learning with convolutional neural networks through the principled lens of version space reduction.
  arXiv  Detail & Related papers  (2020-06-22T17:40:03Z)
• Explainable Deep Classification Models for Domain Generalization [94.43131722655617]
  Explanations are defined as regions of visual evidence upon which a deep classification network makes a decision. Our training strategy enforces a periodic saliency-based feedback to encourage the model to focus on the image regions that directly correspond to the ground-truth object.
  arXiv  Detail & Related papers  (2020-03-13T22:22:15Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.