Impact影响力
Industry collaborations, deployments, open-source reach, and award-winning work from FreedomAI Lab.FreedomAI Lab 的产业合作、真实落地、开源影响力与获奖成果。
Impact at a Glance影响力概览
FreedomAI turns research into deployed systems, open-source resources, industry collaborations, student startups, and recognized papers. The impact highlights below are abstracted from the Shenzhen Key Laboratory of Cross-Modal Cognitive Computing introduction, NSFC materials, and award certificates. FreedomAI 将研究推进到真实系统、开源资源、产业合作、学生创业和获奖论文。下面的影响力信息来自深圳市跨模态认知计算重点实验室介绍、NSFC 材料和获奖证书的整理提炼。
Worldwide Impact全球影响力
- HuatuoGPT intelligent triage and pre-consultation systems have been applied in 11 public hospitals and 186 community health centers in Longgang, Shenzhen, serving about 5 million residents, supporting 300K+ service visits and 520K+ interactions, with reported triage accuracy of 95.4%.HuatuoGPT 智能分诊与预问诊系统已应用于深圳龙岗 11 家公立医院和 186 个社区健康中心，覆盖约 500 万居民，支持 30 万+服务人次和 52 万+交互次数，报告分诊准确率为 95.4%。
- Apollo expands medical LLMs to 50 languages, making medical AI more accessible globally.Apollo 将医学大模型扩展到 50 种语言，让医疗 AI 更容易服务全球用户。
- The lab's medical AI and foundation-model work has been cited or covered by outlets including Nature, Financial Times, and CCTV.实验室的医疗 AI 与基础模型工作曾被 Nature、Financial Times、CCTV 等关注或报道。
GitHub ImpactGitHub 开源影响力
Research code that downstream builders can reuse让下游开发者可以复用的研究代码
The FreedomIntelligence GitHub organization is the engineering backbone for our papers and deployments. It collects public code for medical LLMs, multilingual LLMs, long-context multimodal models, benchmarks, social simulation, mobile agents, and optimization reasoning. FreedomIntelligence GitHub 组织是我们论文、模型和真实部署背后的工程入口，集中开放医疗大模型、多语言大模型、长上下文多模态模型、评测基准、社会模拟、移动智能体和运筹推理等方向的代码。
Hugging Face ImpactHugging Face 模型与数据影响力
Open weights and datasets for the wider AI community面向全球 AI 社区开放模型权重与数据
The FreedomIntelligence Hugging Face hub turns research releases into reusable model weights, datasets, and benchmark resources. It supports medical reasoning, medical vision-language modeling, multilingual medical AI, Arabic LLM localization, dialogue data, and evaluation datasets. FreedomIntelligence Hugging Face 主页把研究成果转化为可直接复用的模型权重、数据集和评测资源，支撑医学推理、医学视觉语言模型、多语言医学 AI、阿拉伯语大模型本地化、对话数据和评测数据等方向。
Awarded Papers and Recognized Work获奖论文与成果认可
Selected best-paper honors and industry recognition from the provided award certificates and NSFC materials. 以下为根据获奖证书和 NSFC 材料整理的代表性最佳论文与产业认可。
Best Paper Award Honorable Mention最佳论文荣誉提名
IRGAN: A Minimax Game for Unifying Generative and Discriminative Information Retrieval Models.IRGAN: A Minimax Game for Unifying Generative and Discriminative Information Retrieval Models。
Best Explainable NLP Paper最佳可解释 NLP 论文
CNM: An Interpretable Complex-valued Network for Matching.CNM: An Interpretable Complex-valued Network for Matching。
Best Paper Award最佳论文奖
Doge Tickets: Uncovering Domain-general Language Models by Playing Lottery Tickets.Doge Tickets: Uncovering Domain-general Language Models by Playing Lottery Tickets。
Financial AI Workshop Best Paper Award金融 AI Workshop 最佳论文奖
TwinMarket: A Scalable Behavioral and Social Simulation for Financial Markets.TwinMarket: A Scalable Behavioral and Social Simulation for Financial Markets。
Industry Recognition for Knowledge Integration知识融合与推理方向产业认可
Huawei recognized a textual knowledge graph plus indexing-enhanced pretraining and inference solution for domain knowledge integration.华为火花奖认可“知识图谱文本化 + 检索增强的预训练及推理”方案在领域知识融合与推理中的价值。
- Additional NSFC-listed recognition includes the NeurIPS 2025 ResponsibleFM Workshop Outstanding Paper Award for work on sparse autoencoder interpretability and utility.NSFC 材料中还记录了稀疏自编码器可解释性与效用相关工作获得 NeurIPS 2025 ResponsibleFM Workshop Outstanding Paper Award。
Industry Collaborations产业合作
We work closely with industry partners across medical AI, multilingual AI, speech, operations research, and agentic systems. 我们与产业伙伴在医疗 AI、多语言 AI、语音、运筹优化和智能体系统等方向紧密合作。
Collaborative Projects合作项目
Student Internships学生实习
