Jung Min LEE (이정민)

M.S. Student, Dept. of Electrical and Computer Engineering
Seoul National University
✉ Email GitHub LinkedIn CV
Jung Min LEE

About

Hi! I'm Jung Min from 🇰🇷. I'm a second-year M.S. student at Cognitive Machine Learning Laboratory, advised by Prof. Jung Woo Lee, in the Department of Electrical and Computer Engineering at Seoul National University (SNU). I am a Research Intern at Physical Intelligence Lab., LG AI Research. I received my B.S. in Physics from SNU. Currently, I am collaborating closely with Li Zhao at Microsoft Research.

Research Keywords

Robot Foundation Models Robot Learning
How to solve real-world problems with generalized intelligence

AI has demonstrated strong performance over the past five years, yet it remains an open question whether AI can reliably solve real-world problems. To address this gap, I believe the most important capability AI must achieve is generalized intelligence.

With this goal in mind, I am currently focusing on data scaling for real-world agents. Data scaling is one of the most effective paths toward generalized intelligence, but it remains challenging in real-world settings. My work aims to develop new methods to scale real-world agent data and, ultimately, enable foundation-model-level generalized intelligence.

News

Jun 2026 I'm going to join LG AI Research's Physical Intelligence Lab as a research intern.
May 2026 Released “Why Latent Actions Fail, and How to Prevent It” on arXiv.
Apr 2026 “MVP-LAM: Learning Action-Centric Latent Action via Cross-Viewpoint Reconstruction” is accepted to ICML 2026!
Feb 2026 Released “MVP-LAM: Learning Action-Centric Latent Action via Cross-Viewpoint Reconstruction” on arXiv.

Education

2025 — Present
M.S., Cognitive Machine Learning Laboratory, Dept. ECE, Seoul National University
Advisor: Prof. Jung Woo Lee  ·  Robot Learning & Reinforcement Learning
2021 — 2025
B.S., Physics and Astronomy & Artificial Intelligence, Seoul National University
Cum Laude
2018 — 2021
Chung-Nam Samsung Academy

Selected Publications

  1. MVP-LAM thumbnail
    MVP-LAM: Learning Action-Centric Latent Action via Cross-Viewpoint Reconstruction
    Jung Min Lee, Dohyeok Lee, Seokhun Ju, Taehyun Cho, Jin Woo Koo, Li Zhao, Sangwoo Hong, Jungwoo Lee
    ICML 2026
    arXiv Code Website
  2. Why Latent Actions Fail thumbnail
    Why Latent Actions Fail, and How to Prevent It
    Jung Min Lee, TaeHyun Cho, Li Zhao, Jungwoo Lee
    arXiv preprint, 2026
    arXiv

Publications

  1. May 2026
    Why Latent Actions Fail, and How to Prevent It
    Jung Min Lee, TaeHyun Cho, Li Zhao, Jungwoo Lee
    arXiv preprint, 2026
    arXiv
  2. Apr 2026
    MVP-LAM: Learning Action-Centric Latent Action via Cross-Viewpoint Reconstruction
    Jung Min Lee, Dohyeok Lee, Seokhun Ju, Taehyun Cho, Jin Woo Koo, Li Zhao, Sangwoo Hong, Jungwoo Lee
    ICML 2026
    arXiv Code Website
  3. Oct 2025
    Learning Generalizable Visuomotor Policy through Dynamics-Alignment
    Dohyeok Lee, Jung Min Lee, Munkyung Kim, Seokhun Ju, Jin Woo Koo, Kyungjae Lee, Dohyeong Kim, TaeHyun Cho, Jungwoo Lee
    arXiv preprint, 2025
    arXiv
  4. Dec 2025
    Viewpoint-Invariant Latent Action Learning from Human Video Demonstrations
    Jung Min Lee, Dohyeok Lee, Jungwoo Lee
    NeurIPS SpaVLE Workshop, 2025
    PDF Website
  5. May 2025
    Dynamics-Aligned Flow Matching Policy for Robot Learning
    Dohyeok Lee, Jung Min Lee, Munkyung Kim, Seokhun Ju, Seungyub Han, Jin Woo Koo, Jungwoo Lee
    CVPR EAI Workshop, 2025
    PDF Website
  6. May 2025
    View-Imagination: Enhancing Visuomotor Control with Adaptive View Synthesis
    Dohyeok Lee, Munkyung Kim, Jung Min Lee, Seungyub Han, Jungwoo Lee
    CVPR EAI Workshop, 2025
    PDF Website

Experience

M.S. Student, Electrical and Computer Engineering

Seoul National University

2025 — Present  ·  Seoul, Korea

Research Intern

LG AI Research — Physical Intelligence Lab.

Jun 2026 — Present  ·  Seoul, Korea

Open Source

IsaacSim-Franka  ⭐ 19