Minsuk Chang

Minsuk Chang is a Research Scientist at Google DeepMind, working at the intersection of Human-Computer Interaction and Machine Learning. His research includes the LLM Comparator visual analytics tool for side-by-side model evaluation, the Concordia framework for studying cooperative social intelligence through generative multi-agent simulations, and work on HyperCLOVA, a billion-scale Korean generative transformer. His current focus is on Gemini post-training to improve goal-driven, multi-step tool use. He earned his Ph.D. in Computer Science from KAIST, advised by Juho Kim, with a dissertation on "Mining Sequential Knowledge for Interaction Design." He was named an ACM AAPI Pioneer in 2023.

• 2024–Present·Google DeepMind·Research Scientist
  • Researching goal-driven, multi-step tool use as part of Gemini post-training.
• 2022–2024·Google·Research Scientist
  • Worked on visual analytics for large language model evaluation (LLM Comparator).
• 2021–2024·KAIST·Adjunct Professor
• 2020–2022·NAVER Corp·Research Scientist
  • Contributed to research on HyperCLOVA, a large-scale Korean generative transformer.

• 2021·KAIST (Korea Advanced Institute of Science and Technology)·Ph.D.
  • Major: Computer Science; Advisor: Juho Kim; Dissertation: "Mining Sequential Knowledge for Interaction Design."