Dongseok Lee
DL
Dongseok Lee

Dongseok Lee

I am currently an M.S. student in Artificial Intelligence at Yonsei University, advised by Prof. Jaehyung Kim at the Machine and Language Learning Lab (ML3). I received my B.S. in Computer Science and Engineering from Sogang University.

My research focuses on improving the capabilities and reliability of large language models, particularly through test-time scaling, knowledge distillation, response verification, and decoding.

Email ↗ GitHub ↗ Google Scholar ↗ CV ↗ ML3 Lab ↗
Test-time scaling Knowledge distillation Post-training Decoding

News

  • 2026 Three papers accepted to EMNLP 2026: ReFeri (Main), LaRA (Findings), and LookBack (Findings).
  • 2025 ReFeri was selected for a spotlight presentation at the ICML 2025 Workshop ES-FoMo-III.

Selected Publications

  1. Training-free LLM Verification via Recycling Few-shot Examples

    EMNLP 2026 · Main

    Dongseok Lee, Jimyung Hong, Dongyoung Kim, Jaehyung Kim

    Also appeared at ICML 2025 Workshop ES-FoMo-III · Spotlight Presentation (14/146, 9.59%).

    Paper Project Code

  2. LaRA: Layer-wise Representation Analysis for Detecting Data Contamination in RL Post-Training

    EMNLP 2026 · Findings

    Minju Gwak, Minseo Kwak, Dongseok Lee, Guijin Son, Alan Ritter, Jaehyung Kim

    Paper Code

  3. LookBack: Where and How to Score LVLM Responses via Visual Reference Usage

    EMNLP 2026 · Findings

    Beomsik Cho*, Jinhyeong Kim*, Dongseok Lee, Jaehyung Kim

    * Equal contribution.

    Paper Code

  4. The Amazing Agent Race: Strong Tool Users, Weak Navigators

    Preprint · 2026

    Zae Myung Kim, Dongseok Lee, Jaehyung Kim, Vipul Raheja, Dongyeop Kang

    Paper Project Code

Experience

  • Machine and Language Learning Lab (ML3), Yonsei University

    Research Student · Advisor: Prof. Jaehyung Kim

    Oct 2024 — Present

  • Qraft Technologies

    Research Intern · AI Trading Solution

    Jul 2024 — Sep 2024

  • Sogang University

    Research Student · Advisor: Prof. HyukJun Lee

    Mar 2024 — Jun 2024

  • Korea Institute of Science and Technology (KIST)

    Research Student · Sensor System Research Center

    Dec 2023 — Feb 2024

Projects

  • LG Electronics

    Mar 2026 — Present

    Scenario-Based Virtual LLM Test User Simulation for Intent Analysis.

  • LG Electronics

    Sep 2025 — Dec 2025

    Automated Evaluation and Validation of LLM-based Intent Analysis.

Education

  • Yonsei University

    M.S. in Artificial Intelligence · Advisor: Prof. Jaehyung Kim

    Sep 2025 — Aug 2027 (Expected)

  • Sogang University

    B.S. in Computer Science and Engineering

    Mar 2019 — Feb 2025

Honors, Service & Teaching

  • Honor2nd Place · SmartFactory Anomaly Detection Competition

    LG AI Research · Mar 2023

  • ServiceConference Reviewer

    International Conference on Learning Representations (ICLR) 2027

  • TeachingTeaching Assistant · Deep Neural Networks (CAS 3201)

    Yonsei University · Fall 2025