Research Papers 论文研究 5h ago Updated 2h ago 更新于 2小时前 49

Faithful by Design: Evaluating and Improving LLM-Generated Clinical Trial Summaries for Multi-Stakeholder Audiences 忠实于设计:评估和改进面向多利益相关者受众的LLM生成临床试验摘要

Introduces a benchmark evaluation framework to measure the faithfulness of LLM-generated clinical trial summaries across three distinct stakeholder audiences. Identifies "Unsupported Claims" as the dominant failure mode for GPT-4o, Claude Sonnet 4.6, and Gemini 2.5 Flash, with a mean annotation score of 1.55/3. Demonstrates that a knowledge-graph-augmented retrieval system yields statistically significant improvements in NLI-based faithfulness scores (p < 0.0001). Reveals model-specific improvem 研究针对LLM在临床摘要生成中的幻觉风险,构建了面向多利益相关者的忠实度评估框架。 基于200项分层临床试验数据,对GPT-4o、Claude Sonnet 4.6和Gemini 2.5 Flash进行基准测试,发现“无依据主张”是主要失败模式。 引入知识图谱增强检索系统,显著提升了模型的忠实度得分(p < 0.0001),不同模型改进路径存在差异。 提出了六维度忠实度标注方案及受众特定的提示模板,为医疗领域LLM应用提供了标准化评估基准。

65
Hot 热度
75
Quality 质量
70
Impact 影响力

Analysis 深度分析

TL;DR

  • Introduces a benchmark evaluation framework to measure the faithfulness of LLM-generated clinical trial summaries across three distinct stakeholder audiences.
  • Identifies "Unsupported Claims" as the dominant failure mode for GPT-4o, Claude Sonnet 4.6, and Gemini 2.5 Flash, with a mean annotation score of 1.55/3.
  • Demonstrates that a knowledge-graph-augmented retrieval system yields statistically significant improvements in NLI-based faithfulness scores (p < 0.0001).
  • Reveals model-specific improvement pathways: GPT-4o reduces contradictions, while Claude Sonnet 4.6 and Gemini 2.5 Flash increase entailment.

Why It Matters

This research addresses critical safety concerns in healthcare AI by providing a rigorous method to evaluate and mitigate hallucinations in high-stakes clinical contexts. It offers actionable insights for developers aiming to integrate LLMs into medical workflows, emphasizing the need for audience-specific evaluation and robust retrieval mechanisms. The findings highlight that generic summarization is insufficient for clinical accuracy, necessitating specialized architectural interventions like knowledge graphs.

Technical Details

  • Evaluation Framework: Utilizes 200 stratified clinical trials from the Aggregate Analysis database, assessed via audience-specific prompt templates and a six-dimension faithfulness annotation schema.
  • Baseline Models: Tests GPT-4o, Claude Sonnet 4.6, and Gemini 2.5 Flash, generating 1,800 summaries scored using a cross-encoder natural language inference (NLI) model.
  • Failure Mode Analysis: "Unsupported Claims" was the primary error source across all models, indicating a systemic issue with factual grounding in current generative architectures.
  • Intervention: Implementation of a knowledge-graph-augmented retrieval system that significantly enhanced faithfulness metrics, proving the efficacy of structured data integration in reducing hallucinations.

Industry Insight

  • Prioritize Retrieval-Augmented Generation (RAG): For clinical applications, standard LLM prompting is inadequate; integrating structured knowledge bases is essential for maintaining factual integrity.
  • Audience-Specific Optimization: Summarization strategies must be tailored to the target audience (providers vs. patients), as faithfulness requirements and failure modes may vary by stakeholder.
  • Model Selection Strategy: Different models exhibit distinct improvement patterns when augmented; GPT-4o benefits from contradiction reduction, while others benefit from entailment increases, suggesting a need for model-specific tuning in production pipelines.

TL;DR

  • 研究针对LLM在临床摘要生成中的幻觉风险,构建了面向多利益相关者的忠实度评估框架。
  • 基于200项分层临床试验数据,对GPT-4o、Claude Sonnet 4.6和Gemini 2.5 Flash进行基准测试,发现“无依据主张”是主要失败模式。
  • 引入知识图谱增强检索系统,显著提升了模型的忠实度得分(p < 0.0001),不同模型改进路径存在差异。
  • 提出了六维度忠实度标注方案及受众特定的提示模板,为医疗领域LLM应用提供了标准化评估基准。

为什么值得看

该研究解决了大语言模型在高风险医疗场景中落地面临的核心信任问题,通过量化评估揭示了主流模型的幻觉特征。其提出的知识图谱增强方案为提升医疗文本生成的准确性提供了可复用的技术路径,对AI在垂直领域的合规应用具有重要参考价值。

技术解析

  • 评估框架与数据集:从Aggregate Analysis数据库抽取200项分层临床试验,构建包含1,800个生成摘要的基准集,采用六维度忠实度标注方案和针对医生、患者、支付方的特定提示模板。
  • 基线模型表现:测试GPT-4o、Claude Sonnet 4.6和Gemini 2.5 Flash,使用交叉编码器自然语言推理(NLI)模型评分,结果显示所有模型的主要错误模式均为“无依据主张”,平均得分为1.55/3。
  • 改进方案:开发了一种知识图谱增强检索系统,通过结构化知识注入减少幻觉。实验显示,该方法使蕴含(entailment)分数提升0.0125,忠实度分数提升0.0130,具有统计学显著性。
  • 模型差异化改进:GPT-4o主要通过减少矛盾来改善表现,而Claude Sonnet 4.6和Gemini 2.5 Flash则主要通过增加正确蕴含来提升分数,表明不同架构对知识增强的响应机制不同。

行业启示

  • 医疗AI需强化事实核查机制:在临床等高风险领域,单纯依赖LLM生成内容风险极高,必须结合外部知识图谱或严格的事实核查模块以确保安全性。
  • 定制化评估标准至关重要:通用基准无法完全反映垂直领域的细微差别,建立针对特定受众(如患者vs医生)和特定任务(如摘要vs诊断)的评估体系是行业标配。
  • 混合架构成为趋势:纯生成式模型在处理高精度要求任务时存在瓶颈,检索增强生成(RAG)特别是结合结构化知识(如知识图谱)的方案将成为提升可靠性的关键手段。

Disclaimer: The above content is generated by AI and is for reference only. 免责声明:以上内容由 AI 生成,仅供参考。

LLM 大模型 Evaluation 评测 Benchmark 基准测试 Healthcare AI 医疗AI Research 科学研究