Liu Leqi
Assistant Professor at UT Austin · Visiting Faculty Researcher at Google
Email: sn.suodc.tuamocce@ unscramble
Office: CBA 6.422
Welcome! My research focuses on understanding and developing capable and controllable machine intelligence. I work on reinforcement learning and post-training for large language models, with a particular interest in settings where the learning signal is imperfect: reward is not easily verifiable, positive trajectories are scarce, and the people supplying feedback have preferences that differ and change over time. This shapes how I approach reasoning, alignment, personalization, and agentic systems. For more information about my research, please visit the HUMAIN Lab.
Current Appointments:
- Assistant Professor (2024 - now) in use-inspired AI at the Department of Information, Risk, and Operations Management, UT Austin
- Affiliated faculty (2025 - now) in Computer Science, UT Austin
- Visiting Faculty Researcher (20% appointment, 2025 - now), Google
Past Experience: I obtained my Ph.D. from the Machine Learning Department at Carnegie Mellon University in 2023, where I was advised by Zachary Lipton. During my Ph.D., I interned at Google DeepMind London in Summer 2022, working on deep reinforcement learning for mechanism design, and at Apple Siri in Summer 2019. I was an Open Philanthropy AI Fellow from 2020 to 2023, and was a postdoctoral researcher at Princeton Language and Intelligence from 2023 to 2024. My Bachelor’s degrees are in Computer Science from Bryn Mawr College and Mathematics from Haverford College. During my undergrad, I spent a wonderful semester at AIT-Budapest in Spring 2016.
Prospective students: Please read our lab website before completing the Google Form to express your interest.