Jeonghoon Shim

Jeonghoon Shim

Ph.D. Student · Graduate School of Data Science
Seoul National University

I am interested in defining and solving problems in the training and evaluation of AI Agents toward real-world deployment. Previously, I've conducted research on dialogue interaction between AI Agents and human users. I am also interested in RL training of AI Agents and visually-grounded Agents (e.g., GUI, Mobile agents).

✉ Email GitHub Google Scholar CV (PDF)
Publications
Non-Collaborative User Simulators
ICLR 2026
Non-Collaborative User Simulators for Tool Agents
Jeonghoon Shim, Woojung Song, Cheyon Jin, Seungwon Koo, and Yohan Jo
  • Identification of non-collaborative user behaviors observed in real-world customer service
  • Proposal of simulation methods that replicate non-collaborative users
  • State-of-the-art and fine-tuned agents are significantly impaired by non-collaborative users
ToolDial
ICLR 2025
ToolDial: Multi-turn Dialogue Generation Method for Tool-Augmented Language Models
Jeonghoon Shim, Gyuhyeon Seo, Cheongsu Lim, and Yohan Jo
  • Syntactic dialogue data generation method using an API graph and agenda-based approach
  • Automatic generation of multi-turn dialogues covering diverse scenarios in quality and quantity
  • State-of-the-art tool agents struggle to perceive exact input parameter states in multi-turn dialogue
Seoul Bike Demand
KST 2022
Analysis of Seoul Public Bike Demand during Night Times by Applying Heckman Selection Models
Jeonghoon Shim and Junho Ko
  • Analysis of shared bicycle usage pattern differences between late-night and daytime hours in Seoul
  • Application of Heckman Selection models on public bike data with geospatial features
  • Current station placement may be misaligned with the needs of late-night users
Education
03/2025 – Present
Ph.D. in Data Science
Seoul National University · Advisor: Yohan Jo
03/2023 – 02/2025
M.S. in Data Science
Seoul National University · Advisor: Yohan Jo
03/2017 – 02/2023
B.S. in Urban Planning and Engineering
Hanyang University · Advisor: Junho Ko
incl. 2-year absence for military service
Industry Collaboration
12/2025 – Present
Yield Analysis LLM Agents for Semiconductor Manufacturing
Samsung Electronics, AI Center
  • Development of a diagnostic framework for yield analysis LLM agents in semiconductor manufacturing
  • Automatic extraction of non-collaborative behavior categories from internal engineer queries
  • Simulation of extracted behaviors and diagnosis of agent weaknesses under such conditions
Teaching
Large Language Models and Conversational AI
Graduate TA · Seoul National University · Spring 2025
Computing for Data Science 1
Graduate TA · Seoul National University · Spring 2024