发表机构
University of Illinois Chicago（伊利诺伊大学芝加哥分校）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究针对基于提示微调的领域自适应模型（DAPF）框架，探究其在痴呆检测任务中的表征优势与标记级解释忠实性不匹配的问题，为痴呆筛查模型的可解释性优化提供了依据。
AI 中文摘要
基于提示的领域自适应模型的口语语言分析是低资源、非侵入性痴呆筛查的有前景方向，但此类模型内部仍不透明。我们研究了基于提示微调的领域自适应模型（DAPF）框架的可解释性，该框架将痴呆检测视为与诊断相关的掩码标记预测任务。我们采用多种探测和分析技术对DAPF及强基线进行解释，发现DAPF取得了最佳整体性能（准确率=0.83，宏F1=0.83），其[MASK]表征可最有效地恢复诊断信息。然而，这种表征优势并未延伸到标记级解释的忠实性。DAPF的归因主要反映语言任务词汇、话语标记和转录伪影，扰动测试显示其影响微弱或为负。这表明其掩码标记接口确定诊断信息，却无法生成忠实的标记级解释。
英文摘要
Spoken-language analysis via prompt-based domain-adaptive models is a promising direction for low-resource, non-invasive dementia screening, but such models remain internally opaque. We study the interpretability of the Domain-Adapted models via Prompt-based Fine-tuning (DAPF) framework, which casts dementia detection as diagnosis-related masked-token prediction. We interpret DAPF and strong baselines using a variety of probing and analysis techniques, finding that DAPF achieved the best overall performance (accuracy=0.83 and macro-F1=0.83) with diagnosis most recoverable from its [MASK] representation. However, this representational advantage did not extend to token-level explanation faithfulness. DAPF attributions primarily reflected language task vocabulary, discourse markers, and transcription artifacts, with perturbation tests showing weak or negative effects. This suggests that its masked-token interface determines diagnosis information without producing faithful token-level explanations.
Comments16 pages, 1 figure, 19 tables. Under review at ACL Rolling Review
