Jongwook Han

han_picture.jpeg

PhD Student @HOLI Lab

Seoul, South Korea

I’m a PhD student at Seoul National University, Graduate School of Data Science, where I am advised by Prof. Yohan Jo. I’m interested in value-alignment in LLMs, and mechanistic interpretability. I received a M.S in Electrical Engineering from Korea Advanced Institute of Science and Technology (KAIST) and a B.S in Integrated Technology from Yonsei University.

Contact

Research Interests

  • Value-alignment in LLMs
  • Mechanistic Interpretability
  • Pluralistic-values in LLMs

Publications

  • Value Portrait: Assessing Language Models’ Values through Psychometrically and Ecologically Valid Items [Paper | Video | Poster]
    Jongwook Han*, Dongmin Choi*, Woojung Song*, Eun-Ju Lee, Yohan Jo
    [ACL] Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics, 2025
  • PVP: An Image Dataset for Personalized Visual Persuasion with Persuasiveness Ratings, Persuasion Strategies, and Viewer Characteristics [Paper | Poster]
    Junseo Kim, Jongwook Han, Dongmin Choi, Jongwook Yoon, Eun-Ju Lee, Yohan Jo
    [ACL] Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics, 2025

Preprints

  • Established Psychometric vs. Ecologically Valid Questionnaires: Rethinking Psychological Assessments in Large Language Models [arXiv]
    Dongmin Choi, Woojung Song, Jongwook Han, Eun-Ju Lee, Yohan Jo
    Under Review
  • Dual Mechanisms of Value Expression: Intrinsic vs. Prompted Values in LLMs [arXiv]
    Jongwook Han*, Jongwon Lim*, Injin Kong, Yohan Jo
    Under Review
  • Quantifying Data Contamination in Psychometric Evaluations of LLMs [arXiv]
    Jongwook Han*, Woojung Song*, Jonggeun Lee*, Yohan Jo
    Under Review
  • Don’t Adapt Small Language Models for Tools; Adapt Tool Schemas to the Models [arXiv]
    Jonggeun Lee*, Woojung Song*, Jongwook Han*, Haesung Pyun, Yohan Jo
    Under Review

Workshops

  • Dual Mechanisms of Value Expression: Decomposing Intrinsic and Prompted Values in Language Models [openreview]
    Jongwook Han*, Jongwon Lim*, Injin Kong, Yohan Jo
    Mechanistic Interpretability Workshop at NeurIPS 2025