I'm a junior at the University of Pennsylvania, where I split my time between research on AI systems and studying for computer science classes. I'm broadly interested in making intelligent systems both smarter and safer.
Between high school and college, I served as a helicopter crew in the South Korean Marines for two years. Now at Penn, I'm on one of the top lightweight rowing teams, which has become a huge part of my college experience.
I've had some great internships at places like IBM Research, NAVER Cloud, and the Center for Axion and Precision Physics Research. You can find my published papers online if you want to see my technical work, though my blog posts are more casual thoughts.
Selected Papers(See all papers)
This paper shows that it is possible to systematically control language model behavior with programmables rules like "if input is about xxx, then refuse." [ICLR 2025 Spotlight]
We argue that language-only models lack understanding of the physical manifestation of language, as demonstrated through a series of tasks called the H-Test. [ACL 2024]
This study presents a synthetic instruction-response generation framework designed to mimic the sequential and orderly nature of human learning. [NAACL 2024]