ConceptSMILE：审计基于概念的可解释人工智能的可信度

中文摘要 AI 辅助

基于概念的可解释人工智能能使模型推理更易理解，但概念级输出并非自动可信。我们引入ConceptSMILE，这是一个基于模型无关扰动的审计框架，用于评估基于概念的解释的可靠性。它扩展了基于扰动的逻辑，从特征或区域级归因到对人类可理解的概念解释的审计。通过对输入区域进行扰动、测量概念响应变化等步骤来评估可靠性。我们在视网膜眼底图像上评估，结果表明不同概念和路径的可靠性不同，ConceptSMILE为评估基于概念的可解释人工智能的可信度提供了独立审计层。

英文摘要

Concept-based explainable artificial intelligence (AI) can make model reasoning more human-understandable, but concept-level outputs are not automatically trustworthy. We introduce ConceptSMILE, a model-agnostic perturbation-based auditing framework for evaluating the reliability of concept-based explanations. Rather than replacing SMILE, ConceptSMILE extends its perturbation-based logic from feature- or region-level attribution to the auditing of human-understandable concept explanations. The framework perturbs input regions, measures concept-response shifts, applies locality weighting, and fits an XGBoost surrogate to approximate local concept behaviour. Reliability is assessed through attribution accuracy, surrogate fidelity, faithfulness, stability, and consistency. We evaluate ConceptSMILE on retinal fundus images by comparing MedSAM-derived visual concepts with VLM-based semantic concepts. Results show that reliability varies across concepts and pathways: MedSAM achieves stronger spatial attribution and the highest surrogate fidelity ($R^2 = 0.8503$, $R_w^2 = 0.8465$), while the VLM pathway shows stronger vessel faithfulness and stronger stability under selected artefact conditions. ConceptSMILE provides an independent audit layer for evaluating the trustworthiness of concept-based XAI.

发表机构

• University of Hull（赫尔大学）
• Newcastle University（纽卡斯尔大学）

机构由 AI 辅助整理，请以论文原文为准。