发表机构
Universidad de Castilla-La Mancha; University of Murcia; Murcian Bio-Health Institute (IMIB-Arrixaca)（卡斯蒂利亚-拉曼恰大学; 穆尔西亚大学; 穆尔西亚生物健康研究所（IMIB-Arrixaca））
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究提出比较可解释性框架审计DeBERTa-v3零样本医学文摘分类，通过五种归因方法比较和Jaccard一致性量化，揭示解释稳定性与预测确定性正相关，并识别三种系统性失败机制，支持多方法联合审计。
AI 中文摘要
本文提出了一个比较可解释性框架，用于审计DeBERTa-v3在医学文摘零样本分类中的表现。该工作解决了可解释人工智能中的分歧问题，即不同的归因方法对相同的输入和预测产生不同的解释。在医学文摘语料库上实现了一个自然语言推理引擎，每个诊断类别使用五个增强假设，每类使用一千个文本的平衡样本。比较了五种解释方法：SHAP和LIME作为模型无关方法，遮挡和输入×梯度作为深度学习特定方法，以及注意力×梯度作为Transformer特定方法。解释通过top-token归因进行标准化，并使用Jaccard指数量化成对一致性。在定义明确的临床领域实现了高预测准确率，而在高语义模糊性下性能下降。解释稳定性直接反映预测确定性，在单值类别中表现出强收敛，在诊断不确定性下显著下降。此外，定性错误审计揭示了三种系统性失败机制：词汇过敏、语义重叠和归因连贯性丧失。结果支持在审计医学文本分类中的Transformer模型时，结合使用多种解释方法和定量一致性指标，并建议优先考虑特定的临床本体论而非宽泛的诊断标签。
英文摘要
A comparative explainability framework is presented to audit DeBERTa-v3 under zero-shot classification of medical abstracts. The work addresses the disagreement problem in Explainable Artificial Intelligence, where different attribution methods produce divergent explanations for the same input and prediction. A natural language inference engine is implemented over the Medical Abstracts corpus with five enriched hypotheses per diagnostic category and a balanced sample of one thousand texts per class. Five explanation methods are compared: SHAP and LIME as model-agnostic approaches, occlusion and Input x Gradient as deep-learning-specific approaches, and Attention x Gradient as a transformer-specific approach. Explanations are standardized through top-token attribution, and pairwise agreement is quantified using the Jaccard index. High predictive accuracy is achieved across well-defined clinical domains, whereas performance degrades under high semantic ambiguity. Explanatory stability directly mirrors predictive certainty, exhibiting strong convergence in univalent categories and a marked drop under diagnostic uncertainty. Furthermore, qualitative error auditing uncovers three systemic failure mechanisms: lexical hypersensitivity, semantic overlap, and loss of attribution coherence. The results support the combined use of several explanation methods and quantitative agreement metrics when auditing transformer-based models in medical text classification, and suggest prioritizing specific clinical ontologies over broad diagnostic labels.
Comments18 pages, 6 figures
