Sojin Lee

M.S. Student
Data Mining Lab,
Dept. of CSE,
Seoul National University
Seoul, South Korea

About Me

I am a M.S. student majoring in Computer Science and Engineering at Seoul National University, advised by Prof. U Kang.

My research interests are large language models (LLM) and model compression, specializing in the study of pruning and quantization for LLMs.

Contact Info.

Education

Publications

* Equal Contribution, † Corresponding Author

Papers

  • A Comprehensive Survey of Compression Algorithms for Language Models
    Seungcheol Park*, Jaehyeon Choi*, Sojin Lee*, and U Kang†
    arXiv 2024 [Paper]

  • Accurate Sublayer Pruning for Large Language Models by Exploiting Latency and Tunability Information
    Seungcheol Park*, Sojin Lee*, Jongjin Kim*, Jinsik Lee, Hyunjik Jo, and U Kang†
    IJCAI 2025 [Paper (TBD) | Code (TBD)]

Research Experience

  • AI Research Project @ Youlchon Foundation (Jun. 2023 - Current)
    "Enhancing the Efficiency of Large Language Models via Extreme Compression"
    Devising a novel model compression algorithm for LLMs

  • LLM Solution @ NAVER Corp. (Jan.-Feb. 2025)
    Developing the pipeline of automatic evaluation for LLMs based on Kubeflow framework

  • PoC Project @ LG AI Research (Feb.-Jun. 2024)
    "Devising a Novel Model Compression Algorithm for Large Language Models"
    Developing sublayer-wise mixed-precision quantization algorithm for LLMs

Academic Services

Teaching Assistant