Xinyi Xie
📍 Palo Alto, California, USA
I am a Ph.D. student in Software Engineering at the University of Arizona, advised by Prof. Bo Liu (RL Turing Award Lab) and Prof. Janet Roveda (Biomedical Engineering), and an AI Engineer at OPPO US Research Center. My research centers on reinforcement learning, agentic AI, AI safety & alignment, and AI for science & healthcare.
I build and evaluate trustworthy LLM agents — from benchmarking and red-teaming to reasoning and post-training. Previously I earned my master's degrees in Economics and Computer Science at Duke University and my bachelor's at Oxford Brookes University. I am always happy to connect with fellow students and researchers — feel free to reach out.
Below are the areas I have worked on:
- AI Agents & Evaluation — designing and benchmarking LLM and multimodal agents, with red-teaming and guardrails that make them reliable enough to deploy.
- Reinforcement Learning & Post-training — alignment, RLHF, reward modeling, and reasoning that shape how large language models behave.
- AI for Science & Healthcare — safe, dependable AI for high-stakes domains such as science, healthcare, and finance.
News
Loading news…
Selected Publications
Loading publications…
Bold = my name. * = equal contribution. See the CV for a full list.
Teaching Experience
- Main Teaching Assistant, Duke University — Quantitative Business Analysis (Summer 2023)
- Teaching Assistant, Duke University — Applied Probability and Statistics (Spring 2022)
Service
Misc.
Outside of research, I spend most of my free time staying active — I love playing golf, and I practise yoga and Pilates regularly. I am also into nutrition and enjoy trying out all kinds of food. Always happy to chat about any of these!