avatar

Jeongeun Lee

M.S. Student
Yonsei University
ljeadec31@yonsei.ac.kr


About Me

I am an 2nd year M.S. student in the Data & Language Intelligence (DLI) Lab at Yonsei University, advised by Prof. Dongha Lee. Before, I received my B.S. in Computer Science from Yonsei University.

If you’re interested in collaborating on research or exploring new ideas together, please feel free to connect with me anytime!


Research Interest

My research focuses on personalized AI that leverages real-world multimodal data, including videos, images, text, and structured signals such as graphs. During my master’s program, I studied (1) personalized evaluation/benchmark, where (M)LLMs reason over multimodal inputs to generate user preference-conditioned rewards, and (2) agentic frameworks, where LLMs simulate real-world user behaviors. More recently, I have become interested in building personalized AI assistants in egocentric and embodied environments with MLLMs, where agents must understand long-term multimodal context and implicit user intent.


News

2026
Jun PIGReward is accepted to ECCV 2026 🇸🇪.
May New preprint is out: POLAR.
2025
Sep CSI is accepted to EMNLP Findings 2025 🇨🇳.
Jul HIPPO-VIDEO is accepted to COLM 2025 🇨🇦.
2024
Jul New preprint is out: CGSP.
Feb Re-Ex is accepted to ICLR 2024 Workshop on Reliable and Responsible Foundation Models.

Publications

*: denotes equal contribution.


  1. ECCV
    Jeongeun Lee, Ryang Heo, Dongha Lee
    European Conference on Computer Vision (ECCV), 2026.

  2. EMNLP Findings
    Tongyoung Kim*, Jeongeun Lee*, Soojin Yoon, Sunghwan Kim, Dongha Lee
    Empirical Methods in Natural Language Processing (EMNLP) Findings, 2025.

  3. COLM
    Jeongeun Lee, Youngjae Yu, Dongha Lee
    Conference on Language Modeling (COLM), 2025.

  4. arXiv
    Jeongeun Lee, Seongku Kang, Won-Yong Shin, Jeongwhan Choi, Noseong Park, Dongha Lee
    arXiv, 2024.

  5. ICLR Workshop
    Juyeon Kim, Jeongeun Lee*, Yoonho Chang*, Chanyeol Choi, Junseong Kim, Jy-yong Sohn
    ICLR Workshop on Reliable and Responsible Foundation Models, 2024.


Powered by Jekyll and Minimal Light theme.