Baichuan-M3：構建臨床問詢模型以實現可靠醫療決策

摘要

我們推出Baichuan-M3——一款醫療增強型大語言模型，旨在將被動問答模式轉變為主動的臨床級決策支持系統。針對現有系統在開放式諮詢中的局限性，Baichuan-M3通過專業化訓練流程模擬醫師的系統性工作流。核心能力包括：（一）主動信息獲取以消除歧義；（二）長視距推理能力，能將零散證據整合為連貫診斷；（三）自適應幻覺抑制機制確保事實可靠性。實證評估顯示，Baichuan-M3在最新推出的HealthBench、HealthBench-Hallu及ScanBench基準測試中取得最優異成果，在臨床問診、健康諮詢與安全性方面顯著超越GPT-5.2。模型已公開於https://huggingface.co/collections/baichuan-inc/baichuan-m3。

English

We introduce Baichuan-M3, a medical-enhanced large language model engineered to shift the paradigm from passive question-answering to active, clinical-grade decision support. Addressing the limitations of existing systems in open-ended consultations, Baichuan-M3 utilizes a specialized training pipeline to model the systematic workflow of a physician. Key capabilities include: (i) proactive information acquisition to resolve ambiguity; (ii) long-horizon reasoning that unifies scattered evidence into coherent diagnoses; and (iii) adaptive hallucination suppression to ensure factual reliability. Empirical evaluations demonstrate that Baichuan-M3 achieves state-of-the-art results on HealthBench, the newly introduced HealthBench-Hallu and ScanBench, significantly outperforming GPT-5.2 in clinical inquiry, advisory and safety. The models are publicly available at https://huggingface.co/collections/baichuan-inc/baichuan-m3.

Baichuan-M3：構建臨床問詢模型以實現可靠醫療決策

Baichuan-M3: Modeling Clinical Inquiry for Reliable Medical Decision-Making

摘要

Support