构建医疗多模态模型幻觉诊断基准，定位错误来源。
ClinHallu: A Benchmark for Diagnosing Stage-Wise Hallucinations in Medical MLLM Reasoning
- 按视觉识别、知识回忆、推理整合三阶段分解推理过程
- 7031个标注实例支持逐阶段错误诊断与干预测试
- 适合医学AI可信性研究者及模型可解释性开发者
构建可信赖的医疗多模态大模型对临床决策支持至关重要。现有医疗幻觉评测多聚焦数据收集，却忽视幻觉在推理过程中的具体来源。我们发现，幻觉成因因样本而异：可能源于视觉误识别、医学知识错误回忆或推理整合缺陷。为实现源级幻觉诊断，我们提出ClinHallu，一个面向医疗多模态大模型推理的阶段化幻觉诊断基准。该基准包含7,031个经验证的实例，每个实例均带有结构化推理轨迹，分解为视觉识别、知识回忆和推理整合三个阶段，并采用阶段替换干预方法，量化修正特定阶段对最终答案的影响。除评估外，我们还证明基于轨迹监督的微调能有效降低各阶段幻觉。ClinHallu为诊断和缓解医疗多模态大模型推理失败提供了细粒度测试平台。基准已开源：https://github.com/alibaba-damo-academy/ClinHallu。
原文摘要 · Abstract (English)
Building trustworthy medical multimodal large language models (MLLMs) is critical for reliable clinical decision support. Existing medical hallucination benchmarks mainly focus on data collection, but often ignore where hallucinations originate within the reasoning process. We find that hallucination sources vary across samples: errors may arise from visual misrecognition, incorrect medical knowledge recall, or flawed reasoning integration. To enable source-level hallucination diagnosis, we introduce ClinHallu, a benchmark for stage-wise hallucination diagnosis in medical MLLM reasoning. ClinHallu contains 7,031 validated instances, where each instance is augmented with a structured reasoning trace decomposed into Visual Recognition, Knowledge Recall, and Reasoning Integration. We also use stage-replacement interventions to measure how correcting specific stages affects the final answer. Beyond evaluation, we show that trace-supervised fine-tuning reduces stage-wise hallucinations. ClinHallu provides a fine-grained hallucination testbed for diagnosing and mitigating reasoning failures in medical MLLMs. The benchmark is publicly available at https://github.com/alibaba-damo-academy/ClinHallu.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。