不用医学微调，也能让大模型答对临床问题。
Do Clinical Question Answering Systems Really Need Specialised Medical Fine Tuning?
- 推理时用轻量向量引导，不改模型权重
- 准确率提升6%，事实一致性增7%，安全错误降50%
- 适合想省成本又怕出错的医疗AI落地场景
临床问答系统日益依赖大语言模型，但普遍认为必须进行医学领域微调。尽管BioBERT、BioGPT、PubMedBERT等专用医学模型流行，却面临覆盖窄、重训成本高、适应性差等现实问题。现有监督微调方法仍强化了‘专业化谬误’——即认为专用模型天然更优。为此，我们提出MEDASSESS-X，一种面向部署的临床问答框架，通过推理时对齐而非微调实现性能稳定。该框架使用轻量级引导向量，使模型激活向医学一致推理方向偏移，无需更新权重或领域重训。实证显示，MEDASSESS-X在通用与专用模型上均实现一致提升：准确率最高+6%，事实一致性+7%，安全错误率降低最多达50%。
原文摘要 · Abstract (English)
Clinical Question-Answering (CQA) industry systems are increasingly rely on Large Language Models (LLMs), yet their deployment is often guided by the assumption that domain-specific fine-tuning is essential. Although specialised medical LLMs such as BioBERT, BioGPT, and PubMedBERT remain popular, they face practical limitations including narrow coverage, high retraining costs, and limited adaptability. Efforts based on Supervised Fine-Tuning (SFT) have attempted to address these assumptions but continue to reinforce what we term the SPECIALISATION FALLACY-the belief that specialised medical LLMs are inherently superior for CQA. To address this assumption, we introduce MEDASSESS-X, a deployment-industry-oriented CQA framework that applies alignment at inference time rather than through SFT. MEDASSESS-X uses lightweight steering vectors to guide model activations toward medically consistent reasoning without updating model weights or requiring domain-specific retraining. This inference-time alignment layer stabilises CQA performance across both general-purpose and specialised medical LLMs, thereby resolving the SPECIALISATION FALLACY. Empirically, MEDASSESS-X delivers consistent gains across all LLM families, improving Accuracy by up to +6%, Factual Consistency by +7%, and reducing Safety Error Rate by as much as 50%.
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。