VISIONLOGIC: From Neuron Activations to Causally Grounded Concept Rules for Vision Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Geng, Chuqin, Jiang, Yuhe, Zhao, Ziyu, Ye, Haolin, Xing, Anqi, Zhang, Li, Si, Xujie
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912921696600064
author Geng, Chuqin
Jiang, Yuhe
Zhao, Ziyu
Ye, Haolin
Xing, Anqi
Zhang, Li
Si, Xujie
author_facet Geng, Chuqin
Jiang, Yuhe
Zhao, Ziyu
Ye, Haolin
Xing, Anqi
Zhang, Li
Si, Xujie
contents While concept-based explanations improve interpretability over local attributions, they often rely on correlational signals and lack causal validation. We introduce VisionLogic, a novel neural-symbolic framework that produces faithful, hierarchical explanations as global logical rules over causally validated concepts. VisionLogic first learns activation thresholds that abstract neuron activations into predicates, then induces class-level logical rules from these predicates. It then grounds predicates to visual concepts via ablation-based causal tests with iterative region refinement, ensuring that discovered concepts correspond to features that are causal for predicate activation. Across different vision architectures such as CNNs and ViTs, it produces interpretable concepts and compact rules that largely preserve the original model's predictive performance. In our large-scale human evaluations, VisionLogic's concept explanations significantly improve participants' understanding of model behavior over prior concept-based methods.
format Preprint
id arxiv_https___arxiv_org_abs_2503_10547
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle VISIONLOGIC: From Neuron Activations to Causally Grounded Concept Rules for Vision Models
Geng, Chuqin
Jiang, Yuhe
Zhao, Ziyu
Ye, Haolin
Xing, Anqi
Zhang, Li
Si, Xujie
Computer Vision and Pattern Recognition
While concept-based explanations improve interpretability over local attributions, they often rely on correlational signals and lack causal validation. We introduce VisionLogic, a novel neural-symbolic framework that produces faithful, hierarchical explanations as global logical rules over causally validated concepts. VisionLogic first learns activation thresholds that abstract neuron activations into predicates, then induces class-level logical rules from these predicates. It then grounds predicates to visual concepts via ablation-based causal tests with iterative region refinement, ensuring that discovered concepts correspond to features that are causal for predicate activation. Across different vision architectures such as CNNs and ViTs, it produces interpretable concepts and compact rules that largely preserve the original model's predictive performance. In our large-scale human evaluations, VisionLogic's concept explanations significantly improve participants' understanding of model behavior over prior concept-based methods.
title VISIONLOGIC: From Neuron Activations to Causally Grounded Concept Rules for Vision Models
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2503.10547