aifollow.news
Back Rohan Paul
Rohan Paul· @rohanpaul_ai · X· · Original publication time AI score47

Ant Group releases medical model LingSante-3.1-flash with available weightsMachine translation

AI introduction

The mixture-of-experts model has 560 billion total parameters and activates 25 billion per token. Its publisher reports that it outperformed GPT-6 Astra, Claude Opus 5.5 and Gemini 4 Argon on HealthBench Professional, HLE and BrowseComp. The model weights are available; the benchmark results are publisher-reported.

Original post · AI translation

@rohanpaul_ai
Translation · refer to the original

蚂蚁集团发布了 LingSante-3.1-flash,这是一款拥有 5600 亿参数的医疗混合专家(MoE)模型;在 HealthBench Professional、HLE、BrowseComp 和 EHR-Complex 上,它的表现超过了 GPT-6 Astra、Claude Opus 5.5 和 Gemini 4 Argon。

HealthBench Professional 涵盖真实临床医生向聊天机器人提出的实际请求。EHR-Complex 测试智能体查询医院病历数据库的能力。

模型权重已在 Hugging Face 上发布,采用 MIT 许可证。

蚂蚁集团从预训练阶段起就混合使用医疗数据和通用数据,将医疗推理、工具使用、搜索以及符合临床医生风格的行为作为独立技能,在经医生和验证器核查的任务上分别给予强化学习奖励进行训练,随后通过多教师同策略蒸馏将这些专门能力融合到一个模型中。

QuoteAnt Ling@AntLingAGI
Translation
今天,我们发布 LingSante-3.1-flash,这是一款开放权重的混合专家模型,总参数量为 5600 亿,每个 token 激活 250 亿参数。它在保留通用能力的同时,整合了多种医疗能力,包括复杂医疗推理、长程工作流程执行,以及与专业行为规范对齐的能力。 LingSante-3.1-flash 超越医疗问答,迈向现实世界中的医疗问题。该模型在 HealthBench Professional(65.72)、HLE(纯文本、使用工具:72.22)和 BrowseComp(提供上下文:94.10)上均居于领先地位,在这 3 项基准测试中都超过了 GPT-6 Astra、Claude Opus 5.5 和 Gemini 4 Argon。LingSante-3.1-flash 在专业长程医疗工作流程中也表现良好,例如 PhysicianBench(57)和 EHR-Complex(74.40)。 这些结果共同表明,LingSante-3.1-flash 是领先的医疗开放权重模型之一。 模型权重现已开放。探索 LingSante,并用它进行开发。 更多内容见下文 ↓
Original

Today, we release LingSante-3.1-flash, an open-weight MoE model with 560B total parameters and 25B activations per token. It unifies diverse medical capabilities while retaining general capabilities, including complex medical reasoning, long-horizon workflow execution, and professional behavior alignment. LingSante-3.1-flash moves beyond medical QA toward real-world healthcare problem. The model leads on HealthBench Professional (65.72), HLE (text-only, with tools: 72.22) and BrowseComp (with context: 94.10), outperforming GPT-6 Astra, Claude Opus 5.5 and Gemini 4 Argon on all three. LingSante-3.1-flash also performs well on professional long-horizon medical workflows, such as PhysicianBench (57) and EHR-Complex (74.40). Together, these results establish LingSante-3.1-flash as the leading open-weight models for healthcare. The weights are available. Explore LingSante and build with it. More below ↓

View this post on X ↗

Source: Rohan Paul · x.com

Model
Found an error? Send a correction