About
I am an M.S. student in Data Science at the University of Michigan, Ann Arbor, with expected graduation in May 2027. My research focuses on temporal point processes (TPPs), agentic AI, and diffusion models. At Michigan, I work with Prof. Raed Al Kontar on graph-aware monitoring and anomaly detection for LLM-based agentic systems, and with Prof. Qing Qu in the DeepThink Lab on diffusion models.
Previously, I worked remotely with Prof. Hongteng Xu at the Gaoling School of Artificial Intelligence, Renmin University of China, developing generative adaptation methods for TPPs that use conditional diffusion to generate context-specific LoRA adapters under distribution shift. I received my B.S. in Physics from Renmin University of China in 2025 and was a visiting student at the University of California, Davis in 2024.
Feel free to contact me if you are interested in related research or potential collaborations.
🔥 News
Our work on generative adaptation of temporal point processes was submitted to AAAI 2027.
I joined Prof. Qing Qu's DeepThink Lab at the University of Michigan to work on diffusion models.
I began collaborating remotely with Prof. Hongteng Xu at the Gaoling School of Artificial Intelligence, Renmin University of China, on generative adaptation for temporal point processes.
I joined Prof. Raed Al Kontar's group as a research assistant, working on reliable LLM agent systems.
I started the M.S. in Data Science program at the University of Michigan, Ann Arbor.
I graduated from Renmin University of China with a bachelor's degree. 🎓
🔬 Research Experience
Graph-Aware Monitoring and Anomaly Detection in LLM Agent Systems
Research Assistant · University of Michigan · Advisor: Prof. Raed Al Kontar
Developing graph-aware monitoring for LLM-based agentic systems using proxy telemetry and terminal outcomes, including two-channel sequential detectors for early anomaly detection and diagnosis.
Understanding Autoregressive and Random-Order Generation
Research Assistant · University of Michigan · Advisor: Prof. Qing Qu
Investigating why autoregressive (AR) and random-order generation perform differently across datasets. This ongoing work develops a theoretical explanation and explores VAR-style training to identify generation orders best suited to AR modeling.
Generative Adaptation of Temporal Point Processes
Research Assistant · Renmin University of China · Advisor: Prof. Hongteng Xu
Proposed GA-TPP for zero-shot adaptation under distribution shift. Conditional diffusion generates context-specific LoRA adapters for a pretrained temporal point process, producing strong out-of-distribution gains on real-world datasets.
Deep Learning for Particle-in-Cell Simulation
Bachelor's Thesis · Renmin University of China · Advisor: Prof. Weimin Wang
Replaced the Poisson solver in a 1D particle-in-cell framework with MLP and CNN-MLP neural field predictors, reaching approximately 10^-3 MSE and reproducing two-stream instability.
💼 Industry Experience
Research Analyst Intern
Guosen Securities · Fixed Income Investment Department
Developed ARIMA/LSTM forecasts and a DQN trading framework with volatility-aware rewards, achieving approximately 5% in-sample and 4% out-of-sample excess returns in the reported experiments.
Investment Banking Winter Intern
Guotai Haitong Securities · Investment Banking Department
Conducted IPO due diligence, analyzed financial and ownership structures, assessed shareholder and competition risks, and contributed to prospectus drafting.
🎓 Education
University of Michigan, Ann Arbor
Department of Statistics · Master's Degree
Renmin University of China
School of Physics · Bachelor's Degree
University of California, Davis
Visiting Student