Seungjong Sun
I am a Ph.D. Candidate in the Department of Applied Artificial Intelligence at Sungkyunkwan University, working in the Data Science and Social Analytics Lab (DSSAL) under Prof. Jang Hyun Kim. I expect to graduate in February 2027 and am on the job market for postdoctoral positions.
Research Interests
I study how large language models imitate human personality, emotion, and interaction — what human values and behaviors they carry, and how those can be read and edited in the model's internals.
- Human Values and Behavior in LLMs. I ask how LLMs reproduce human personality, interaction, and social behavior, and how these are encoded in their internal representations.
- Interpretability and Model Control. I use activation steering and model diffing to locate the internal directions behind these behaviors, and model merging to align models without additional training.
I also work on AI safety, developing evaluation benchmarks for LLM safeguards in sarcasm understanding and jailbreak robustness.
News
Selected Publications
Conference
Sarcasm Subtype-Specific Reasoning in Dialogue with Multimodal Cues Using Large Language Models
Kiss up, Kick down: Exploring Behavioral Changes in Multi-modal Large Language Models with Assigned Visual Personas
Preprint
Random Silicon Sampling: Simulating Human Sub-Population Opinion Using a Large Language Model Based on Group-Level Demographic Information
Full publication list in my CV and on Google Scholar.
Teaching
Instructor
Teaching Assistant
Education
Research Experience
Talks & Service
Invited Talk
Reviewing