CV
Agent systems and evaluation, statistical modeling, and scientific machine learning. Download the two-page CV for research contributions and selected publications.
Education & academic positions
- Mar 2026 – present Postdoctoral Researcher Halıcıoğlu Data Science Institute, UC San Diego
- Oct 2020 – Feb 2026 Ph.D. in Particle Physics University of California, San Diego
- Aug 2016 – Jun 2020 B.S. in Physics Kuang Yaming Honors School, Nanjing University
Selected roles
Projects →- Feb 2026 – present Sole architect and developer, SIDERIUS Resource-aware autonomous research, composable agent capabilities, and scientific evaluation
- Jul 2026 – present Founder and Coordinator, SciTra Trajectory and action-space representation for scientific agents — approximately 100 scientists across 20+ institutions
- Oct 2020 – Feb 2026 Statistical inference, XENONnT Collaboration One of the corresponding authors on the solar-neutrino measurement
- Jan 2024 – Feb 2026 Maintainer, strax and straxen Open-source data frameworks for 200+ collaborators
Technical skills
- Programming & ML Python, PyTorch, NumPy, Pandas, SciPy, scikit-learn, C/C++, Bash, Linux, Git
- Agents & evaluation LLM agent systems, multi-agent orchestration, agent evaluation, benchmark design, action-space and trajectory instrumentation, autonomous experimental loops
- Statistics Bayesian and maximum-likelihood inference, uncertainty propagation, sensitivity estimation, hypothesis testing, probabilistic modeling
- Systems Pydantic, FastAPI, pytest, GitHub Actions, SLURM/HPC, ETL pipelines, resource monitoring and performance optimisation
Service
Talks →- Mar 2026 Session chair, APS Global Physics Summit 2026 Machine Learning Methods at the Intersection of Nuclear and Neutrino Physics (APR-R87)