Hyunseok Lee

I am a second-year Ph.D. student at KAIST, advised by Prof. Jinwoo Shin, and a Research Intern at NAVER Cloud. My research interests lie in building intelligence that is self-aware and safe. To this end, I focus on developing Large Language Models (LLMs) that can reason, make decisions (i.e., exhibit agentic behaviors), and ensure safety. I am also broadly interested in the continual pretraining of LLMs and their multilingual capabilities.

Topics of interest: LLM, LLM Reasoning, LLM-based Agents, LLM Safety

* denotes equal contribution

ReGUIDE: Data Efficient GUI Grounding via Spatial Reasoning and Search

Hyunseok Lee, Jeonghoon Kim, Beomjun Kim, Jihoon Tack, Chansong Jo, Jaehong Lee, Cheonbok Park, Sookyo In, Jinwoo Shin, Kang Min Yoo
paper | code

ReVISE: Learning to Refine at Test-Time via Intrinsic Self-Verification

Hyunseok Lee*, Seunghyuk Oh*, Jaehyung Kim, Jinwoo Shin, Jihoon Tack
ICML 2025
paper | code

ReMoDetect: Reward Models Recognize Aligned LLM's Generations

Hyunseok Lee*, Jihoon Tack*, Jinwoo Shin
NeurIPS 2024
Qualcomm Innovation Fellowship Korea 2024
paper | code | slide | poster | project page | demo

Ph.D. in Artificial Intelligence
Korea Advanced Institute of Science and Technology (KAIST), Mar. 2024 - Current
Advisor: Prof. Jinwoo Shin

B.S. in Electrical Engineering and Computer Science (double major)
Korea Advanced Institute of Science and Technology (KAIST), Mar. 2018 - Feb. 2024

Qualcomm Innovation Fellowship Korea 2024
Research Intern, Naver Cloud, Feb. 2025 - Present
Research Intern, Omnious AI, Jun. 2023 - Dec. 2023
1st Place in Korean LLM Leaderboard, hyunseoki/ko-en-llama2-13b, Oct. 2023
Conference Reviewer, NeurIPS
Workshop Reviewer, Reasoning and Planning for LLMs@ICLR