I am currently a Postdoctoral Fellow at Stanford Institute for Human-Centered Artificial Intelligence (HAI) working with Alex ‘Sandy’ Pentland, Diyi Yang and Erik Brynjolfsson. I am affiliated with the Stanford Digital Economy Lab and the Stanford NLP group. I obtained my PhD from School of Information at the University of Michigan advised by David Jurgens. I also worked with Jun Li at Ross School of Business. Outside of academia, I have interned at Bloomberg, Snapchat, Nokia Bell Labs, and Tencent.

I will join the School of Information at UT-Austin as an Assistant Professor in Fall 2026 and will start a new lab on Human-Centered AI Systems. I will be hiring PhD, Master’s, undergraduate, and visiting students. Please stay tuned for more details!

Research Interests

My research focuses on Large Language Models, Human-Computer Interaction and Computational Social Science. I am currently interested in the following research directions:

1. Social NLP: How to evaluate and improve LLMs’ capabilities to understand and interact with humans in social contexts.

2. Human-AI Interaction: How to design and evaluate AI systems for different social scenarios.

3. Computational Social Science: How to understand human behaviors and social dynamics using large-scale data and computational methods.

Opening

I am always looking for talented students and collaborators from all backgrounds for both research and impactful open source projects. Feel free to shoot me an email if you are interested in my research, but please understand if I am not able to respond to every email.

Awards & Recognition

  • Best Student Paper Award, ACM Conference on Equity and Access in Algorithms, Mechanisms, and Optimization (EAAMO)
  • Best Demo Paper Award, AAAI Conference on Human Computation and Crowdsourcing (HCOMP)
  • Honorable Mention Award, International Conference on Computational Social Science (IC2S2)
  • Best Paper Award, Workshop on Social Influence in Conversations

Software

I am the core developer of POTATO, an open-source data labeling system for effective and responsive human computation.