Baichuan Intelligence Unveils M4 Model, Enabling Doctor-Like Active Diagnosis
The medical field has seen significant advances with the latest AI applications. Baidu Intelligent and a Tsinghua University research team have jointly launched Baichuan-M4, a next-generation medical-enhanced large language model. Leveraging its strong logical reasoning and clinical knowledge, the model achieved top scores in the authoritative HealthBench medical evaluation, leading in three sub-charts as well as the Hard and Professional categories, showcasing medical expertise that exceeds GPT-5.5.
The core advancement of Baichuan-M4 lies in a fundamental shift in its interaction model. Traditional AI-assisted diagnosis is often limited to passive question-and-answer exchanges, whereas Baichuan-M4 simulates a real doctor’s diagnostic reasoning, enabling proactive questioning. It deeply analyzes patient-provided information, independently identifies clinical clues, and asks targeted follow-up questions, greatly enhancing both the speed and accuracy of diagnosis, positioning AI as a reliable clinical decision-making partner.

To enhance diagnostic continuity, M4 introduces the "Full Course Memory" feature. This technology deeply integrates a patient's historical medical records, multi-round consultation logs, trends in lab results, and past medication responses, ensuring the model maintains access to the complete medical history across multiple conversations without needing to start over or repeat information. In long-context clinical memory assessments, the model performed exceptionally well, showing significant improvements over its predecessor.
For clinical safety and accuracy, Baichuan-M4 introduces "Evidence Anchoring" technology. Every medical conclusion generated by the model is precisely linked to specific paragraphs in authoritative medical papers or clinical guidelines, rather than just general literature citations. Data shows its evidence-based reference accuracy reaches 90.0, significantly outperforming other large models in the rigor of medical decision-making.
Related article
Microsoft Upgrades Copilot Notebooks With Task Delegation, Office Editing, and Interactive Reports
Microsoft is rolling out key updates to Copilot Notebooks, including task delegation, in-place Office document editing, and interactive reports. This shift transforms the tool from a simple AI organizer into a comprehensive smart workspace for execut
Meta, Broadcom and TSMC to Mass-Production Self-Developed AI Chip Iris in September
Meta is teaming up with Broadcom and TSMC to build a custom AI processor dubbed "Iris," targeting mass production by September 2026. As part of Meta's MTIA initiative, the chip has cleared initial testing in just six weeks. Designed to optimize ranki
Google AI Overviews Gain Ground in Search Results
According to Similarweb’s 2026 Generative AI Landscape report, Google’s AI-generated summaries appeared in 43% of US searches, a significant rise from 15% recorded a year prior.While AI Overviews present generated summaries within standard search res
Related Special Topic Recommendations
Comments (0)
0/500
The medical field has seen significant advances with the latest AI applications. Baidu Intelligent and a Tsinghua University research team have jointly launched Baichuan-M4, a next-generation medical-enhanced large language model. Leveraging its strong logical reasoning and clinical knowledge, the model achieved top scores in the authoritative HealthBench medical evaluation, leading in three sub-charts as well as the Hard and Professional categories, showcasing medical expertise that exceeds GPT-5.5.
The core advancement of Baichuan-M4 lies in a fundamental shift in its interaction model. Traditional AI-assisted diagnosis is often limited to passive question-and-answer exchanges, whereas Baichuan-M4 simulates a real doctor’s diagnostic reasoning, enabling proactive questioning. It deeply analyzes patient-provided information, independently identifies clinical clues, and asks targeted follow-up questions, greatly enhancing both the speed and accuracy of diagnosis, positioning AI as a reliable clinical decision-making partner.

To enhance diagnostic continuity, M4 introduces the "Full Course Memory" feature. This technology deeply integrates a patient's historical medical records, multi-round consultation logs, trends in lab results, and past medication responses, ensuring the model maintains access to the complete medical history across multiple conversations without needing to start over or repeat information. In long-context clinical memory assessments, the model performed exceptionally well, showing significant improvements over its predecessor.
For clinical safety and accuracy, Baichuan-M4 introduces "Evidence Anchoring" technology. Every medical conclusion generated by the model is precisely linked to specific paragraphs in authoritative medical papers or clinical guidelines, rather than just general literature citations. Data shows its evidence-based reference accuracy reaches 90.0, significantly outperforming other large models in the rigor of medical decision-making.
Microsoft Upgrades Copilot Notebooks With Task Delegation, Office Editing, and Interactive Reports
Microsoft is rolling out key updates to Copilot Notebooks, including task delegation, in-place Office document editing, and interactive reports. This shift transforms the tool from a simple AI organizer into a comprehensive smart workspace for execut
Meta, Broadcom and TSMC to Mass-Production Self-Developed AI Chip Iris in September
Meta is teaming up with Broadcom and TSMC to build a custom AI processor dubbed "Iris," targeting mass production by September 2026. As part of Meta's MTIA initiative, the chip has cleared initial testing in just six weeks. Designed to optimize ranki
Google AI Overviews Gain Ground in Search Results
According to Similarweb’s 2026 Generative AI Landscape report, Google’s AI-generated summaries appeared in 43% of US searches, a significant rise from 15% recorded a year prior.While AI Overviews present generated summaries within standard search res





Home






