JUNGTAEK KIM

Research Associate
at the University of Wisconsin–Madison

jungtaek.kim [at] wisc [dot] edu

Google Scholar · GitHub · X · LinkedIn · Curriculum Vitae

HOME · DETAILED PROFILE · ARCHIVES · PUBLICATIONS: COMPLETE LIST

NEWS

[February 2026] My collaborative paper "TAPE: Tool-Guided Adaptive Planning and Constrained Execution in Language Model Agents" with Jongwon Jeong and Prof. Kangwook Lee has been released at arXiv.
[February 2026] My collaborative paper "GP-4DGS: Probabilistic Analysis of 4D Gaussian Splattings for Monocular Video Reconstruction via Variational Gaussian Processes" with Mijeong Kim and Prof. Bohyung Han has been accepted at CVPR-2026.
[January 2026] Happy new year!
[December 2025] My collaborative paper "ReJump: A Tree-Jump Representation for Analyzing and Improving LLM Reasoning" with Yuchen Zeng, Shuibai Zhang, Wonjun Kang, Prof. Kangwook Lee, and others has been released at arXiv.
[May 2025] My two papers "Density Ratio Estimation-based Bayesian Optimization with Semi-Supervised Learning" and "VersaPRM: Multi-Domain Process Reward Model via Synthetic Reasoning Data" have been accepted at ICML-2025.

ABOUT ME

Jungtaek Kim is a research associate at the University of Wisconsin–Madison, working with Prof. Kangwook Lee. Previously, he was a postdoctoral associate at the University of Pittsburgh, working with Prof. Paul W. Leu, Prof. Satish Iyengar, Prof. Lucas Mentch, and Prof. Oliver Hinder, and a postdoctoral researcher at POSTECH in the group of Prof. Minsu Cho. He received a B.S. in Mechanical Engineering and Computer Science and Engineering from POSTECH in 2015 and a Ph.D. in Computer Science and Engineering from POSTECH in 2022, under the supervision of Prof. Seungjin Choi and Prof. Minsu Cho. During his Ph.D. program, he interned at Vector Institute and SigOpt (acquired by Intel). His internship at Vector Institute was carried out under the supervision of Prof. Graham W. Taylor, and he worked with Dr. Michael McCourt at SigOpt. He presented his work as a first author or co-author at top-tier machine learning conferences such as NeurIPS, ICML, AISTATS, UAI, and ICLR. He has also served as a program committee member or reviewer for several machine learning conferences, including NeurIPS, ICML, AISTATS, UAI, and ICLR. His main research interests are statistical machine learning, Bayesian optimization, large language models, and sequential assembly.

CURRENT RESEARCH INTERESTS

Statistical Machine Learning
Bayesian Optimization
Large Language Models
Sequential Assembly

WORK-IN-PROGRESS PAPERS

Jongwon Jeong, Jungtaek Kim, and Kangwook Lee. TAPE: Tool-Guided Adaptive Planning and Constrained Execution in Language Model Agents.
Yuchen Zeng, Shuibai Zhang, Wonjun Kang, Shutong Wu, Lynnix Zou, Ying Fan, Heeju Kim, Ziqian Lin, Jungtaek Kim, Hyung Il Koo, Dimitris Papailiopoulos, and Kangwook Lee. ReJump: A Tree-Jump Representation for Analyzing and Improving LLM Reasoning.
Jungtaek Kim. Beyond Regrets: Geometric Metrics for Bayesian Optimization.

ELSEWHERE

Academic Profile: Google Scholar, DBLP, Semantic Scholar, ORCID
Social Network: LinkedIn, X
Development: GitHub

Last updated: February 24, 2026