Who are you, and what do you work on?
I'm Changyoon. I study how large language models can teach without taking over. My research sits at the intersection of NLP and education, focused on learner agency and understanding in LLM tutoring.
About

I'm a Ph.D. student at KAIST's School of Computing, advised by Alice Oh, where I work on the educational applications of large language models and natural language processing. My research integrates NLP, AI, and education to understand learners better, and to build models and tutoring systems that help learners develop genuine understanding, supporting their reasoning rather than replacing it.

A recurring theme in my work is learner agency: designing LLM-based tutors that keep the learner in the driver's seat, prompting reflection, surfacing misconceptions, and adapting to what a student actually understands. I develop methods to capture learner understanding and context from learner–educator interactions, and personalized pedagogical approaches that bring the most educational benefit to learners.

LLMs for education Learner agency NLP & understanding Human–AI interaction
Research focus

Tutoring that builds understanding — not dependency

A few questions that guide my work:

How do we measure and preserve learner agency in a tutoring dialogue?

What makes an LLM's pedagogical moves trustworthy and adaptive to an individual student?

When does an AI tutor deepen a learner's understanding, and when does it just hand over the answer?

Recent news All news