HeeJun Kim

Heejun Kim

📍 Gwangju Institute of Science and Technology (GIST)

Major: Electrical Engineering & Computer Science
Minor: AI & Mathematics

Mail | CV | GitHub | Google Scholar

I am a senior undergraduate student at GIST with a deep fascination for AGI, LLM reasoning, and efficient AI inference. My research focuses on understanding how intelligent systems can improve their reasoning capabilities through interaction and adaptation.

Currently, I am conducting research on LLM uncertainty-based agentic memory at the Learning and Adaptation Algorithm Lab under Prof. Taesup Kim's supervision. I am also working on LLM with reinforcement learning at the Data Science Lab under Prof. Sundong Kim's supervision.

Through the Korea-Germany Junior Research Fellowship, I had the opportunity to work as a visiting researcher at Forschungszentrum Jülich's INM-7 institute directed by Prof. Simon B. Eickhoff, where I investigated the intersection of neuroscience and machine learning through brain connectivity analysis.

Education

Gwangju Institute of Science and Technology (GIST)

February 2020 - Present

  • B.S Major: Electrical Engineering and Computer Science
  • GPA: 4.21/4.50
University of California, Berkeley (UC Berkeley)

June 2024 - August 2024

  • Summer Session, GPA: 4.0/4.0
  • Coursework: Introduction to Artificial Intelligence, Principles & Techniques of Data Science

Research Experience

Learning and Adaptation Algorithm Lab - Research Assistant

Seoul National University (SNU) | Jan 2026 - Present

Supervised by Prof. Taesup Kim

  • Conducting research on Agentic Memory Using LLM Uncertainty Estimation
Data Science Lab - Research Assistant

Gwangju Institute of Science and Technology (GIST) | Feb 2025 - Present, Jan 2024 - June 2024

Supervised by Prof. Sundong Kim

  • Developed architecture applying generative models to reinforcement learning for Abstract Reasoning Corpus (ARC)
  • Researched test-time adaptation with reinforcement learning for self-improving reasoning capabilities of large language models
Forschungszentrum Jülich - Visiting Researcher

September 2024 - February 2025

Host Institute: INM-7, directed by Prof. Simon B. Eickhoff and contributed to Mathematical Neuroscience Group led by Dr. Oleksandr Popovych

  • Developed preprocessing pipeline for multi-modal MRI human connectome data using High-Performance Computing (HPC) and HyperThreading technology
  • Korea-Germany Junior Research Fellowship 2024 organized by Max Planck Korea
  • Funded by National Research Foundation of Korea (NRF): 12,000,000KRW (8,197EUR)

Publications

Efficient Epistemic Uncertainty Estimation for Large Language Models via Knowledge Distillation

Seonghyeon Park, Jewon Yeom, Jaewon Sok, Jeongjae Park, Heejun Kim, Taesup Kim

arXiv:2602.01956 | Paper

Diffusion-Based Offline RL for Improved Decision-Making in Augmented ARC Task

Yunho Kim, Jaehyun Park, Heejun Kim, Sejin Kim, Byung-Jun Lee, Sundong Kim

arXiv:2410.11324 | Paper

Honors & Awards

  • Government Scholarship, GIST (Mar 2020 - Present)
  • Government STEM Scholarship (Mar 2020 - Present)
  • Dean's List, College of Information and Computing, GIST (Spring 2025, Fall 2025)
  • Korea-Germany Junior Research Fellowship Program 2024 (Sep 2024 - Feb 2025)
  • GIST International Internship Support Program (Sep 2024 - Feb 2025)
  • Hyundai AI Challenge - Top 25/330 teams (Sep 2023 - Oct 2023)