Zongzhe Xu
PhD Student, Computer Science
University of California, Los Angeles
Google Scholar / LinkedIn / CV
About
I am a first-year Computer Science PhD student at UCLA, working with Prof. Yuzhe Yang in the Health Intelligence Lab (HAIL). Before UCLA, I completed my M.S. in Machine Learning at Carnegie Mellon University, where I worked with Prof. Ameet Talwalkar and Prof. Mikhail Khodak in the Sage Lab. Earlier, I earned my B.S. in Computer Science and Mathematics (summa cum laude) at Washington University in St. Louis, where I did clinical research with Dr. Peinan Zhao at the OBGYN department of the WashU School of Medicine.
Outside of academia, I spent time at Datadog as a Data Scientist Intern contributing to a time-series foundation model.
What I Work On
I build machine learning systems for health, with a current focus on physiological and wearable biosignals such as polysomnography and motion data. A person's health is only partially captured by any single measurement, and I am interested in models that can learn useful, transferable representations from these noisy, multimodal, and often incomplete observations. My current directions include:
- Foundation models for biosignals & wearable health data
- Multimodal learning & language interfaces for health
- Sleep and physiological signal understanding
More broadly, I care about designing models whose architectures and objectives are tailored to the structure of real measurements that remain useful, interpretable, and transferable across clinical settings. I am always happy to chat about collaborations in these areas.
News
- [Apr. 2026]Homepage is up! Always looking for collaborations in AI for health.
- [2026]SleepLM: Natural-Language Intelligence for Human Sleep is on arXiv!
- [2026]OSF: On Pre-training and Scaling of Sleep Foundation Models is on arXiv!
- [Sep. 2025]Our work on time-series foundation models, This Time is Different: An Observability Perspective on Time Series Foundation Models, accepted at NeurIPS 2025!
- [Aug. 2025]Started my PhD at UCLA in the HAIL Lab with Prof. Yuzhe Yang!
- [Jan. 2025]Joined Datadog as a Data Scientist Intern working on TOTO.
- [Jan. 2025]Specialized Foundation Models Struggle to Beat Supervised Baselines accepted at ICLR 2025!
- [Jan. 2025]V2X-DG: Domain Generalization for Vehicle-to-Everything Cooperative Perception accepted at ICRA 2025!
- [Dec. 2024]Graduated from CMU with an M.S. in Machine Learning.
Publications
A few recent works I am most excited about. See my Google Scholar for the complete list. (* indicates equal contribution.)
Education
- [Aug. 2025 – Present] Ph.D. in Computer Science, University of California, Los Angeles.
- [Aug. 2023 – Dec. 2024] M.S. in Machine Learning, Carnegie Mellon University.
- [Sep. 2019 – May 2023] B.S. in Computer Science & Mathematics (summa cum laude, minors in Bioinformatics and Biology), Washington University in St. Louis.
Teaching
- Teaching Assistant, CSE 412 Introduction to Artificial Intelligence, Washington University in St. Louis, Sep. 2021 – Jun. 2022.
- Teaching Assistant, CSE 417 Machine Learning Theory, Washington University in St. Louis, Sep. 2021 – Jun. 2022.
A Few Things About Me
Away from the keyboard, I have been falling deep into bouldering and climbing, and I am steadily working my way into more hiking and skiing. Long before I picked up a lab notebook, I played competitive badminton — I hold a National Second-Class Athlete certification in China. If you want to talk research, share a climbing gym recommendation, or rally a few points, drop me a line.