About
机构
Yueqian Labs is a research and engineering institute focused on post-LLM infrastructure. We build systems for constrained intelligence: bounded inference, verified outputs, local execution.
The institute operates on the premise that the most consequential AI systems will be those that function under defined constraints rather than those that maximize unconstrained capability. Our work sits at the intersection of verification, privacy, and deployment under resource limits.
We do not build chatbots. We do not optimize engagement metrics. We do not pursue artificial general intelligence as a product category.
月前实验室是一家专注于后LLM基础设施的研究与工程机构。我们构建面向约束智能的系统：有界推断、已验证输出、本地执行。
本机构的运作基于这样的前提：最具影响力的AI系统将是那些在定义约束下运行的系统，而非那些追求无约束能力最大化的系统。我们的工作位于验证、隐私和资源受限部署的交汇处。
我们不构建聊天机器人。我们不优化互动指标。我们不将通用人工智能作为产品类别追求。
lab note: yl/ab/001Constraint-first design
约束优先设计
Systems are designed around their constraints, not retrofitted with guardrails.
Verification ships with the system
验证随系统交付
No system deploys without its evaluation harness.
Local by default
默认本地运行
Inference runs locally unless there is a defined reason to do otherwise.
Privacy is structural
隐私是结构性的
Privacy is an architecture decision, not a policy overlay.
Bounded scope
有界范围
Every engagement, every system, every claim has defined bounds.
Institutional restraint
机构克制
We do not make claims we cannot verify. We do not ship systems we cannot evaluate.
Research Systems
研究系统
Core research programs across all five domains.
Edge Inference Group
边缘推断组
Deployment and thermal profiling for constrained hardware.
Local Compute Program
本地计算项目
Airgapped environments, secure runtimes, local retrieval.
Quantum Workflows Unit
量子工作流单元
Hybrid classical-quantum workflow design.
Nano Research Division
纳米研究部
Nano interfaces, materials-adjacent compute, signal processing.
Ruohan Chen
Founder & Director
Former research scientist at Google DeepMind. PhD in Computer Science from Tsinghua University. Published work on inference verification and constrained orchestration systems. Founded Yueqian Labs in 2024 to pursue post-LLM infrastructure research independently.
创始人兼总监。前谷歌DeepMind研究科学家。清华大学计算机科学博士。
Maya Vasquez
Head of Edge Inference
Previously led thermal-aware deployment at Qualcomm AI Research. Specializes in offline inference under hardware constraints. Leads the Edge Inference Group (yl-div-002) and the thermal profiling lab.
边缘推断负责人。前高通AI研究部热感知部署负责人。
Kenji Tanaka
Head of Quantum Workflows
PhD in Quantum Computing from Caltech. Previously at IBM Quantum, working on hybrid classical-quantum optimization. Leads bounded sampling research and decoherence-aware scheduling programs.
量子工作流负责人。加州理工学院量子计算博士。前IBM量子研究员。
Lian Zhang
Head of Research Systems
Previously at Anthropic, working on AI governance and audit systems. MS in Information Security from UC Berkeley. Leads governance trace format development and verification budget research.
研究系统负责人。前Anthropic AI治理与审计系统工程师。加州大学伯克利分校信息安全硕士。
Prof. Samuel Okonkwo
Professor of Computer Engineering, UCLA. Secure runtimes and data sovereignty.
Dr. Akiko Nishida
Former Director of AI Safety, RIKEN. Verification systems and bounded inference.
Daniel Osei, PhD
CTO, Constrained Systems Inc. Edge deployment and thermal profiling.
This site serves as the public-facing interface for Yueqian Labs. Content is selectively disclosed. Not all programs are listed. Not all divisions are visible.
本站作为月前实验室的对外界面。内容为选择性披露。并非所有项目均已列出。并非所有部门均可见。
lab note: yl/ab/sites/002All external contact is processed through the controlled intake system. We do not accept unsolicited engagement outside defined channels. Responses are issued within stated timelines. Inquiries that fall outside current program scope will be acknowledged and filed.
所有对外联系均通过受控入口系统处理。我们不接受定义渠道之外的主动联系。回复将在规定时限内发出。超出当前项目范围的问询将被确认并归档。
contact intake → lab note: yl/ab/comms/002