Biography

Hello. I am a Ph.D. student in Electrical and Computer Engineering (ECE) at the University of Wisconsin-Madison, advised by Prof. Kangwook Lee and Prof. Robert Nowak. I am currently in IBM Research (AI Foundation Model Team) as a research intern in summer 2026.

Previously, I was an Applied AI Research Scientist at Deep Learning Div. in KRAFTON. Also, I was an AI Research Scientist at Applied AI Lab, NLP Center in NCSOFT. I received B.S. and M.S. degrees in Electrical Engineering (EE) from Korea Advanced Institute of Science and Technology (KAIST) in 2018 and 2020, respectively, where I was advised by Prof. Sae-Young Chung.

A detailed curriculum vitae (CV) is available here.

Research Interest: LLM / VLM Agent

My research interests focus on Large Language Model (LLM) / Vision language Model (VLM) Agents that are practically usable in real-world applications.

To this end, I pursue two directions: (1) improving agent capability under real-usage cases by identifying key limitations of current agents and developing empirical and theoretical methods to address them, and (2) improving agent efficiency by building computationally efficient agents, often based on small language models (sLMs), that close the performance gap with large models at lower compute costs.

From an application perspective, I am also interested in applying LLM/VLM agents to domains such as Embodied AI, Game AI, and so on.

Research Keywords

• LLM / VLM Agents; Agentic Reasoning; Small Language Models (sLMs); Knowledge Distillation

• Embodied AI; Game AI; Programming

News

• (June. 2026) 🎉  Our paper "Exploration and Exploitation Are Measurable in Large Language Model Agents" has been accepted at ICML 2026 Workhop on Failure Modes in Agentic AI (FAGEN)!

• (May. 2026) ✨ I start a Research intern at IBM Research in Summer 2026!

• (Apr. 2026) 🎉 Our paper "How to Correctly Report LLM-as-a-Judge Evaluations" has been accepted at ICML 2026! See you in Seoul 🏝️!

  A full list of my past updates is available here.

Selected Publications

• Exploration and Exploitation Are Measurable in Large Language Model Agents [Code] [Project page]
  Jaden Park*, Jungtaek Kim*, Jongwon Jeong*, Robert D. Nowak, Kangwook Lee, Yongjae Lee (*: Equally Contributed)
  ICML 2026 Workshop on Failure Modes in Agentic AI (FAGEN)

• How to Correctly Report LLM-as-a-Judge Evaluations [Code]
  Chungpa Lee, Thomas Zeng, Jongwon Jeong, Jy-yong Sohn, Kangwook Lee
  International Conference on Machine Learning (ICML), 2026

• TAPE: Tool-Guided Adaptive Planning and Constrained Execution in Language Model Agents [Code]
  Jongwon Jeong, Jungtaek Kim, Kangwook Lee
  ICLR 2026 Workshop on Agentic AI in the Wild, 2026

• T1: Tool-integrated Verification for Test-time Compute Scaling in Small Language Models
  Minki Kang*, Jongwon Jeong*, Jaewoong Cho (*: Equally Contributed)
  International Conference on Learning Representations (ICLR), 2026

• Distilling LLM Agent into Small Models with Retrieval and Code Tools  [Code]
  Minki Kang, Jongwon Jeong, Seanie Lee, Jaewoong Cho, Sung Ju Hwang
  Neural Information Processing Systems (NeurIPS), 2025 (Spotlight)

  A full list of my work is available here.