I study language models, and my long-term goal is to build human-inspired AI.
Lately I’m interested in looped transformers, diffusion language models, and latent reasoning.
Education
-
2025 – present
M.S. in Computer Science and Engineering, POSTECH
Advisor: Prof. Hwanjo Yu, co-advisor: Prof. Sangwoo Mo
-
2020 – 2025
B.S. in Industrial Management and Engineering, POSTECH
Publications
Work experience
-
Jun 2023 – Aug 2023
Bagelcode
Research assistant intern
Market research, game genre analysis, and KPI-based validation of marketing strategy.
-
Jul 2022 – Dec 2022
Penta Security
DevOps intern
Automated web UI testing with Selenium, converted a hardware-installed product into a Docker image, and wrote CI/CD scripts for deploying containers from an in-house release server.
Awards
-
2024
Gold Prize, 4th UNIST–KAIST–POSTECH AI & Data Science
Built a FinanceRAG model for U.S. financial disclosure analysis and Q&A, third in retrieval performance; proposed a translation and summarization service for Japanese stock information, highest score in the service idea category.
-
2024
Honorable mention, DB Finance & Economics Contest
Report: “Single-stock ETF as an alternative to short selling: impact on the underlying asset market.”
-
2023
Argus Labs gaming track, Interchain Hackathon
Onchain Visual Novel, a text-based relationship simulation using the GPT-4 API, at Korea Blockchain Week 2023. Led the planning and LLM design; built the game server on World Engine and the web client.
-
2023
Finalist, WorldQuant International Quant Championship
Developed and backtested stock price forecasting models on WorldQuant BRAIN; team placed in the top eight in Korea.
Off hours
Outside of research, I spend most of my time watching films and listening to music.