TU Berlin & Huawei Munich Research Center
Hi! I'm an industrial PhD researcher at TU Berlin and Huawei Munich Research Center, working with Prof. Stefan Schmid and Dr. Yong Li.
I work on LLM agent security and privacy-preserving machine learning. I'm interested in how agents can use tools safely, stay within the permissions users give them, and keep sensitive data private. My work also covers secure LLM inference and private retrieval-augmented generation.
Before my PhD, I studied at KTH Royal Institute of Technology and Southeast University. I also worked on security at Huawei 2012 Labs and did my master's thesis at Ericsson in Stockholm. Outside research, I enjoy traveling and taking photos.
Warning
Problem: The current name of your GitHub Pages repository ("Solution: Please consider renaming the repository to "
http://".
However, if the current repository name is intended, you can ignore this message by removing "{% include widgets/debug_repo_name.html %}" in index.html.
Action required
Problem: The current root path of this site is "baseurl ("_config.yml.
Solution: Please set the
baseurl in _config.yml to "Education
-
Technische Universität BerlinJoint program with Huawei Munich Research CenterPh.D. in AI SecurityAug. 2024 - Present
-
KTH Royal Institute of TechnologySecurity and Mobile ComputingM.Sc.Sep. 2021 - Jun. 2023
-
Southeast UniversitySchool of Information Science and EngineeringB.Eng. in Information EngineeringSep. 2018 - Jun. 2022
Experience
-
Huawei Munich Research CenterAI Security and Privacy-Preserving ComputingIndustrial PhD Research InternAug. 2024 - Present
-
Huawei Technologies Co., Ltd.2012 LabsCybersecurity and Privacy EngineerSep. 2023 - Jul. 2024
-
Ericsson StockholmMaster Thesis ProjectMaster Thesis StudentJan. 2023 - Jun. 2023
Honors & Awards
-
Outstanding PhD Student, Huawei Munich Research Center2024, 2025
-
Fully Funded Industrial PhD Scholarship, Huawei Munich Research Center2024 - 2028
-
Undergraduate Overseas Joint Training Scholarship, Southeast University2021
Academic Service
-
Reviewer, ICLR2027
-
Reviewer, ECML PKDD2026
News
Selected Publications (view all )
SecureClaw: Clawing Back Control of LLM Agents
Yuhan Ma, Stefan Schmid
Conference on Neural Information Processing Systems (NeurIPS) 2026 Accepted
A security architecture for tool-using LLM agents that confines sensitive plaintext at the read boundary and authorizes external actions through a trusted executor.
SecureClaw: Clawing Back Control of LLM Agents
Yuhan Ma, Stefan Schmid
Conference on Neural Information Processing Systems (NeurIPS) 2026 Accepted
A security architecture for tool-using LLM agents that confines sensitive plaintext at the read boundary and authorizes external actions through a trusted executor.
FuseFSS: Efficient Secure LLM Inference with Function Secret Sharing
Yuhan Ma, Yong Li, Stefan Schmid
International Conference on Machine Learning (ICML) 2026 Accepted
A function-secret-sharing based framework for efficient two-server secure LLM inference, targeting fixed-point comparisons, interval lookup, low-degree polynomial evaluation, and predicate-dependent computation.
FuseFSS: Efficient Secure LLM Inference with Function Secret Sharing
Yuhan Ma, Yong Li, Stefan Schmid
International Conference on Machine Learning (ICML) 2026 Accepted
A function-secret-sharing based framework for efficient two-server secure LLM inference, targeting fixed-point comparisons, interval lookup, low-degree polynomial evaluation, and predicate-dependent computation.