从归因到回避：基于注意力的无训练审计用于临床摘要
From Attribution to Abstention: Training-Free Attention-Based Auditing for Clinical Summarization
AI总结：
本文提出ClinTrace框架，通过解码器注意力权重提取临床有用信号，实现无训练的注意力基于审计，提升临床摘要的透明性和可靠性，显著优于基线方法。
摘要：
在临床摘要中部署多模态大语言模型（MLLM）需要不仅流畅生成，还要透明地说明每个陈述的来源，并提供一种标记缺乏证据支持陈述的机制。我们提出了ClinTrace，一个无训练框架，从每个基于transformer的MLLM在生成过程中产生的解码器注意力权重中提取两个临床有用的信号：（i）细粒度的源归因，将每个输出句子与支持文本跨度或图像链接，以及（ii）每句话的 groundedness 分数，识别 poorly supported 声称作为候选幻觉。这两个信号均从相同的注意力张量中在单次传递中提取，无需重新训练、无需辅助模型和额外的推理成本。我们评估了两个临床摘要任务：医生-患者对话摘要（CliConSummation）和放射学报告摘要（MIMIC-CXR），使用通用的MLLM（Qwen3-8B）和医疗微调模型（HuatuoGPT-Vision-7B）。对于源归因，ClinTrace在放射学摘要中达到超过92%的文本F1，在对话摘要中达到88%，显著优于基于嵌入和自归因基线。对于幻觉检测，groundedness分数在医疗微调模型中达到0.77 AUROC：与基于嵌入的置信度相比，零额外成本竞争，并启用一种回避机制，通过 withholding 最少的20%输出给临床医生审查，将忠实度从61.7%提高到72.6%。值得注意的是，医疗微调显著提高了基于注意力的幻觉检测的可靠性，表明领域适应产生更具语义结构的注意力模式，更适用于自审计。
英文摘要：
Deploying multimodal large language models (MLLMs) for clinical summarization demands not only fluent generation but also transparency about where each statement originates-and a mechanism to flag when statements lack evidential support. We present ClinTrace, a training-free framework that extracts two clinically useful signals from the decoder attention weights that every transformer-based MLLM already produces during generation: (i) fine-grained source attributions linking each output sentence to supporting text spans or images, and (ii) per-sentence groundedness scores that identify poorly supported claims as candidate hallucinations. Both signals are derived from the same attention tensors in a single pass, requiring no retraining, no auxiliary models, and no additional inference cost. We evaluate on two clinical summarization tasks: doctor-patient dialogue summarization (CliConSummation) and radiology report summarization (MIMIC-CXR) using a general-purpose MLLM (Qwen3-8B) and a medical-finetuned model (HuatuoGPT-Vision-7B). For source attribution, ClinTrace achieves over 92% text F1 on radiology and 88% on dialogue summarization, substantially outperforming embedding-based and self-attribution baselines. For hallucination detection, groundedness scores achieve 0.77 AUROC with the medical-finetuned model: competitive with embedding-based confidence at zero additional cost, and enable an abstention mechanism that improves faithfulness from 61.7% to 72.6% by withholding the least: grounded 20% of output for clinician review. Notably, medical finetuning substantially improves the reliability of attention-based hallucination detection, suggesting that domain adaptation produces more semantically structured attention patterns amenable to self-auditing.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
