About Me

Hi~ I’m Hantao Lou, an undergraduate student(‘26) majoring in Artificial Intelligence at Yuanpei College, Peking University.

Currently I’m doing research at Center of AI Safety and Governance, Institute of Aritificial Intelligence, Peking University. I’m fortunate to be advised by Prof. Yaodong Yang. Also I’m a scholar at MATS program, under the mentorship of Evan Hubinger doing research about developmental interpretability and model audit.

At Peking University, I’m a member and the monitor of Tong Class, an pilot class in Aritificial Intelligence.

My research interests mainly include Alignment Algorithms, Mechanistic Interpretability(or for short, mech interp), and other potentially scalable methods. My research questions are:

1. How can the findings from mechanistic interpretability be effectively integrated into practical applications, including the alignment process?
2. How can the fundamental nature of intelligence be uncovered through the interpretation of various models that exhibit intelligent behavior?

I’m still learning mech interp, so it is quite possible that I may revise these questions in the future.

I’m open to collaborations and discussions. Feel free to reach out to me via email: hantaolou.htlou at gmail dot com