Leekyeung Han

I am a Ph.D. student in the Department of Computer Science and Engineering at Korea University, advised by Prof. Paul Hongsuck Seo at the Multimodal Interactive Intelligence Lab (MIIL).

My research is on dialog-based embodied AI and large language / vision-language models (LLM/VLM) — building agents that perceive their surroundings, hold a multi-turn conversation with a human partner, and act on what was said. Most recently I have worked on the DialNav benchmark for multi-turn dialog navigation and on scaling it up through automatic embodied dialog augmentation.

Education

  • Korea University, Seoul, Korea — Ph.D. Student, Computer Science and Engineering
    Mar. 2024 – Present · Advisor: Paul Hongsuck Seo (MIIL)
  • Korea University, Seoul, Korea — B.S. in Computer Science
    Mar. 2015 – Feb. 2019

Experience

  • Backend Engineer — SK Telecom, Seoul, Korea
    Jan. 2019 – Present

Publications

  • Revisiting Task Vectors: A Structural Analysis of Their Limits
    Leekyeung Han, Hyeseung Ji, Sangwon Jung, Seungwoo Lyu, Paul Hongsuck Seo
    Findings of EMNLP 2026

  • Advancing DialNav through Automatic Embodied Dialog Augmentation
    Leekyeung Han, Sangwon Jung, Hyunji Min, Jinseong Jeong, Minyoung Kim, Paul Hongsuck Seo
    EMNLP 2026 · arXiv

  • GOAT: A Training Framework for Goal-Oriented Agent with Tools
    Hyunji Min, Sangwon Jung, Junyoung Sung, Dosung Lee, Leekyeung Han, Paul Hongsuck Seo
    Findings of ACL 2026 · arXiv

  • DialNav: Multi-turn Dialog Navigation with a Remote Guide
    Leekyeung Han, Hyunji Min, Gyeom Hwangbo, Jonghyun Choi, Paul Hongsuck Seo
    ICCV 2025 · arXiv

Organizing & Service

Awards

  • Best Paper Award, Korea University, 2025 — for DialNav: Multi-turn Dialog Navigation with a Remote Guide