Machine Learning & Intelligent Control Lab.

Principal Investigator

증명사진_한승열

Seungyul Han

Associate Professor

AI Graduate School (AIGS) & Department of Electrical Engineering (EE)

Ulsan National Institute of Science and Technology (UNIST)

50 Unist-gil, Ulsan, 44919, South Korea

[Google Scholar]  [Curriculum Vitae]

Contact:

  • Tel         +82-52-217-3455
  • Office    Room. 501-2, Building 106, UNIST

Biography

Seungyul Han is an Associate Professor in the Artificial Intelligence Graduate School (AIGS) and the Department of Electrical Engineering (EE) at the Ulsan National Institute of Science and Technology (UNIST). He received his B.S. degree (double major in Mathematical Sciences) and M.S. degree in Electrical Engineering from the Korea Advanced Institute of Science and Technology (KAIST), Daejeon, South Korea, in 2013 and 2016, respectively, and his Ph.D. degree in Electrical Engineering from KAIST. Prior to joining UNIST, he was a postdoctoral researcher at the Institute of Information Electronics at KAIST. He served as an assistant professor at UNIST until 2026 and is currently an associate professor at UNIST. His research interests include reinforcement learning, multi-agent systems, LLM-based decision making, machine learning & deep learning, and RL-based control systems.

Research Areas

Our laboratory advances machine learning (ML) and reinforcement learning (RL) grounded in rigorous mathematical principles. We explore the foundations and frontiers of next-generation AI by examining the complementary integration of large language models (LLMs) and RL, developing RL algorithms for physical intelligence and embodied agents, and advancing fundamental research in learning-based control and decision making. We also investigate diverse real-world domains where RL can deliver transformative impact. Through this integrated research program, we aim to establish the core control and reasoning capabilities that will underpin future AI-driven systems. The four research directions described above are summarized in the figure below:

LLM demo
LLM-based Decision Making

LLM-based Decision Making

  • Language-conditioned policies
  • Instruction / task following with LLMs
  • LLM-guided RL planning
  • Cross-domain imitation via language
physical demo
Physical Intelligence

Physical Intelligence

  • Imitation / behavior cloning
  • Domain adaptation & sim-to-real
  • Intelligent control systems
  • Signal / sensor-based learning
core demo
Core Reinforcement Learning

Core Reinforcement Learning

  • Offline / batch RL
  • Multi-agent RL
  • Meta reinforcement learning
  • Robust & safe learning
  • Statistical / theoretical RL
applied demo
Multi-Agent Systems

Applied Reinforcement Learning

  • Industrial automation
  • Autonomous / multi-agent systems
  • Process / logistics optimization
  • Real-world ML applications

News

[2026/09] Our 4 papers have been accepted to NeurIPS 2026. Congratulations!

[2026/09] Jeongha Park and Myoungseok Han joined our lab !

[2026/08] Paper “MARBO: Relational Belief Grounding for LLM Agents in Social Deduction Games” has been accepted to EMNLP 2026. Congratulations!

[2026/05] Our 2 papers have been accepted to ICML 2026. Congratulations!

[2026/03] Prof. Han (PI) has been promoted to Associate Professor.

[2026/01] Our 4 papers have been accepted to ICLR 2026. Congratulations!

[2025/12] Paper “Generalized Per-Agent Advantage Estimation for Multi-Agent Policy Optimization” has been accepted to AAMAS 2026 (Best paper nominated). Congratulations!

[2025/09] Gwanwoo Choi, Mingu Kang, and Sangwoo Bang joined our lab !

[2025/08] Project PI: “End-to-End 대규모 자율군집 제어를 위한 LLM기반 다중 에이전트 강화학습 기술 개발” (NRF 국가아젠다 기초연구, 25.09 – 28.08, 총 연구비 4.8억)

[2025/05] Our 2 papers have been accepted to ICML 2025. Congratulations!

[2025/03] Junhyun Bae and Yechan Whang joined our lab !

[2024/09] Paper “Exclusively Penalized Q-learning for Offline Reinforcement Learning” has been accepted to NeurIPS 2024 as a spotlight paper. Congratulations!

[2022/04] Project PI: “자율 드론 실용화를 위한 목적지향 강화학습 핵심기술 개발” (IITP 사람중심인공지능 핵심원천기술개발, 22.04 – 26.12, 총 연구비 47.5억)