Saved in:
Bibliographic Details
Main Authors: Zhao, Delong, Huang, Qiang, Yan, Di, Sun, Yiqun, Yu, Jun
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2511.22170
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917108457144320
author Zhao, Delong
Huang, Qiang
Yan, Di
Sun, Yiqun
Yu, Jun
author_facet Zhao, Delong
Huang, Qiang
Yan, Di
Sun, Yiqun
Yu, Jun
contents Concept Bottleneck Models (CBMs) enhance interpretability by introducing a layer of human-understandable concepts between inputs and predictions. While recent methods automate concept generation using Large Language Models (LLMs) and Vision-Language Models (VLMs), they still face three fundamental challenges: poor visual grounding, concept redundancy, and the absence of principled metrics to balance predictive accuracy and concept compactness. We introduce PS-CBM, a Partially Shared CBM framework that addresses these limitations through three core components: (1) a multimodal concept generator that integrates LLM-derived semantics with exemplar-based visual cues; (2) a Partially Shared Concept Strategy that merges concepts based on activation patterns to balance specificity and compactness; and (3) Concept-Efficient Accuracy (CEA), a post-hoc metric that jointly captures both predictive accuracy and concept compactness. Extensive experiments on eleven diverse datasets show that PS-CBM consistently outperforms state-of-the-art CBMs, improving classification accuracy by 1.0%-7.4% and CEA by 2.0%-9.5%, while requiring significantly fewer concepts. These results underscore PS-CBM's effectiveness in achieving both high accuracy and strong interpretability.
format Preprint
id arxiv_https___arxiv_org_abs_2511_22170
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Partially Shared Concept Bottleneck Models
Zhao, Delong
Huang, Qiang
Yan, Di
Sun, Yiqun
Yu, Jun
Computer Vision and Pattern Recognition
Concept Bottleneck Models (CBMs) enhance interpretability by introducing a layer of human-understandable concepts between inputs and predictions. While recent methods automate concept generation using Large Language Models (LLMs) and Vision-Language Models (VLMs), they still face three fundamental challenges: poor visual grounding, concept redundancy, and the absence of principled metrics to balance predictive accuracy and concept compactness. We introduce PS-CBM, a Partially Shared CBM framework that addresses these limitations through three core components: (1) a multimodal concept generator that integrates LLM-derived semantics with exemplar-based visual cues; (2) a Partially Shared Concept Strategy that merges concepts based on activation patterns to balance specificity and compactness; and (3) Concept-Efficient Accuracy (CEA), a post-hoc metric that jointly captures both predictive accuracy and concept compactness. Extensive experiments on eleven diverse datasets show that PS-CBM consistently outperforms state-of-the-art CBMs, improving classification accuracy by 1.0%-7.4% and CEA by 2.0%-9.5%, while requiring significantly fewer concepts. These results underscore PS-CBM's effectiveness in achieving both high accuracy and strong interpretability.
title Partially Shared Concept Bottleneck Models
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2511.22170