Qinan Yu 喻启楠
Hi! I am Qinan, a second-year CS Ph.D. student at Stanford. I am advised by Chris Potts and Carlos Guestrin. My Ph.D. is generously supported by the Stanford Graduate Fellowship.
My research focuses on interpretability and trustworthiness. I am interested in understanding, explaining, and controlling models’ reasoning processes. I aim to build reliable and faithful explanations of how models reason to help humans make decisions and learn from these models. Moreover, I am excited to design faithful, reliable, and scalable interpretability methods for real-world applications.
Before Stanford, I received a concurrent B.S. in Math-CS and M.A. in CS from Brown, where I was fortunate to be advised by Ellie Pavlick.
News
-
May 2025Our paper "Improved Representation Steering for Language Models" is now available on arXiv.
-
Jan 2025Our paper "The Same But Different: Structural Similarities and Differences in Multilingual Language Modeling" is accepted into ICLR 2025.
-
Sep 2024Start my Ph.D. at Stanford.
See my publications page for a complete list of papers.