Jongmin Lee
I'm a postdoc at UC Berkeley, advised by Pieter Abbeel. I received my PhD from KAIST, where I was fortunate to be advised by Kee-Eung Kim.
E-mail: jongmin.lee012 [at] gmail dot com / jongmin.lee [at] berkeley dot edu
Education
2017. 03. - 2022. 02: PhD, School of Computing, KAIST, Korea (Advisor: Kee-Eung Kim)
Thesis: Algorithms for Safe Reinforcement Learning
2015. 03. - 2017. 02.: MS, School of Computing, KAIST, Korea (Advisor: Kee-Eung Kim)
Thesis: Constrained Bayesian Reinforcement Learning via Approximate Linear Programming
2009. 03. - 2014. 02.: BS, Department of Computer Science and Engineering, Seoul National University, Korea
Experience
May 2022 - Current: Postdoc at UC Berkeley (Advisor: Pieter Abbeel)
Apr 2021 - Aug 2021: Research Scientist Intern at DeepMind (Host: Arthur Guez)
Publications
International
[C25] Mitigating Covariate Shift in Behavioral Cloning via Robust Distribution Correction Estimation
Seokin Seo, Byung-Jun Lee, Jongmin Lee, HyeongJoo Hwang, Hongseok Yang, Kee-Eung Kim
[C24] ROIDICE: Offline Return on Investment Maximization for Efficient Decision Making
Woosung Kim, Hayeong Lee, Jongmin Lee, Byung-Jun Lee
[C23] Body Transformer: Leveraging Robot Embodiment for Policy Learning [paper] [website]
Carmelo Sferrazza, Dun-Ming Huang, Fangchen Liu, Jongmin Lee, Pieter Abbeel
[C22] Kernel Metric Learning for In-Sample Off-Policy Evaluation of Deterministic RL Policies [paper]
Haanvid Lee, Tri Wahyu Guntara, Jongmin Lee, Yung-Kyun Noh, Kee-Eung Kim
ICLR 2024 (spotlight)
[C21] AlberDICE: Addressing Out-Of-Distribution Joint Actions in Offline Multi-Agent RL via Alternating Stationary Distribution Correction Estimation [paper] [code]
Daiki E. Matsunaga*, Jongmin Lee*, Jaeseok Yoon, Stefanos Leonardos, Pieter Abbeel, Kee-Eung Kim (*: equal contribution)
[C20] SafeDICE: Offline Safe Imitation Learning with Non-Preferred Demonstrations [paper]
Youngsoo Jang, Geon-Hyeong Kim, Jongmin Lee, Sungryull Sohn, Byoungjip Kim, Honglak Lee, Moontae Lee
[C19] Tempo Adaptation in Non-stationary Reinforcement Learning [paper]
Hyunin Lee, Yuhao Ding, Jongmin Lee, Ming Jin, Javad Lavaei, Somayeh Sojoudi
[C18] LobsDICE: Offline Imitation Learning from Observation via Stationary Distribution Correction Estimation [paper]
Geon-Hyeong Kim*, Jongmin Lee*, Youngsoo Jang, Hongseok Yang, Kee-Eung Kim (*: equal contribution)
[C17] Local Metric Learning for Off-Policy Evaluation in Contextual Bandits with Continuous Actions
Haanvid Lee, Jongmin Lee, Yunseon Choi, Wonseok Jeon, Byung-Jun Lee, Yung-Kyun Noh, Kee-Eung Kim
[C16] COptiDICE: Offline Constrained Reinforcement Learning via Stationary Distribution Correction Estimation [paper] [code]
Jongmin Lee, Cosmin Paduraru, Daniel J. Mankowitz, Nicolas Heess, Doina Precup, Kee-Eung Kim, Arthur Guez
ICLR 2022 (spotlight)
[C15] DemoDICE: Offline Imitation Learning with Supplementary Imperfect Demonstrations [paper] [code]
Geon-Hyeong Kim, Seokin Seo, Jongmin Lee, Wonseok Jeon, HyeongJoo Hwang, Hongseok Yang, Kee-Eung Kim
[C14] GPT-Critic: Offline Reinforcement Learning for End-to-End Task-Oriented Dialogue Systems [paper]
Youngsoo Jang, Jongmin Lee, Kee-Eung Kim
[C13,W4] OptiDICE: Offline Policy Optimization via Stationary Distribution Correction Estimation [paper] [code]
Jongmin Lee*, Wonseok Jeon*, Byung-Jun Lee, Joelle Pineau, Kee-Eung Kim (*: equal contribution)
[C12] Representation Balancing Offline Model-based Reinforcement Learning [paper] [code]
Byung-Jun Lee, Jongmin Lee, Kee-Eung Kim
[C11] Monte-Carlo Planning and Learning with Language Action Value Estimates [paper] [code]
Youngsoo Jang, Seokin Seo, Jongmin Lee, Kee-Eung Kim
[C10] Reinforcement Learning for Control with Multiple Frequencies [paper] [code]
Jongmin Lee, Byung-Jun Lee, Kee-Eung Kim
[C9] Batch Reinforcement Learning with Hyperparameter Gradients [paper] [code]
Byung-Jun Lee*, Jongmin Lee*, Peter Vrancx, Dongho Kim, Kee-Eung Kim (*: equal contribution)
[C8] Monte-Carlo Tree Search in Continuous Action Spaces with Value Gradients [paper]
Jongmin Lee, Wonseok Jeon, Geon-Hyeong Kim, Kee-Eung Kim
[C7,W4] Bayes-Adaptive Monte-Carlo Planning and Learning for Goal-Oriented Dialogues [paper]
Youngsoo Jang, Jongmin Lee, Kee-Eung Kim
[C6] Trust Region Sequential Variational Inference [paper]
Geon-Hyeong Kim, Youngsoo Jang, Jongmin Lee, Wonseok Jeon, Hongseok Yang, and Kee-Eung Kim
[C5] PyOpenDial: A Python-based Domain-Independent Toolkit for Developing Spoken Dialogue Systems with Probabilistic Rules [paper] [code]
Youngsoo Jang*, Jongmin Lee*, Jaeyoung Park*, Kyeng-Hun Lee, Pierre Lison, and Kee-Eung Kim (*: equal contribution)
[C4] Monte-Carlo Tree Search for Constrained POMDPs [paper] [code]
Jongmin Lee, Geon-Hyeong Kim, Pascal Poupart, and Kee-Eung Kim
[W3] Monte-Carlo Tree Search for Constrained MDPs [paper]
Jongmin Lee, Geon-Hyeong Kim, Pascal Poupart, and Kee-Eung Kim
[J1] Layered Behavior Modeling via Combining Descriptive and Prescriptive Approaches: a Case Study of Infantry Company Engagement [paper]
Jang Won Bae, Junseok Lee, Do-Hyung Kim, Kanghoon Lee, Jongmin Lee, Kee-Eung Kim and Il-Chul Moon
IEEE Transactions on System, Man, and Cybernetics: Systems, 2018
[C3,W2] Constrained Bayesian Reinforcement Learning via Approximate Linear Programming [paper]
Jongmin Lee, Youngsoo Jang, Pascal Poupart, and Kee-Eung Kim
Scaling-Up Reinforcement Learning Workshop at ECML PKDD (SURL), 2017
[C2] Hierarchically-partitioned Gaussian Process Approximation [paper]
Byung-Jun Lee, Jongmin Lee, and Kee-Eung Kim
[W1] Multi-View Automatic Lip-Reading using Neural Network [paper]
Daehyun Lee, Jongmin Lee, and Kee-Eung Kim
ACCV Workshop on Multi-view Lip-reading/Audio-visual Challenges, 2016
[C1] Bayesian Reinforcement Learning with Behavioral Feedback [paper]
Teakgyu Hong, Jongmin Lee, Kee-Eung Kim, Pedro A. Ortega, and Daniel Lee
Domestic
A Study on Efficient Multi-Task Offline Model-based Reinforcement Learning
Geon-Hyeong Kim, Youngsoo Jang, Jongmin Lee, and Kee-Eung Kim
한국소프트웨어종합학술대회, 2021
A Study on Application of Efficient Lifelong Learning Algorithm to Model-based Reinforcement Learning
Byung-Jun Lee, Jongmin Lee, Yunseon Choi, Youngsoo Jang, and Kee-Eung Kim
한국소프트웨어종합학술대회, 2020
A Study on Monte-Carlo Tree Search in Continuous Action Spaces
Jongmin Lee, Geon-Hyeong Kim, and Kee-Eung Kim
한국통신학회 하계종합학술발표회 논문집, 2019
Case Studies on Planning and Learning for Large-Scale CGFs with POMDPs through Counterfire and Mechanized Infantry Scenarios
Jongmin Lee, Jungpyo Hong, Jaeyoung Park, Kanghoon Lee, Kee-Eung Kim, Il-Chul Moon, and Jae-Hyun Park
KIISE Transactions on Computing Practices, 2017
A Case Study on Planning and Learning for Large-Scale CGFs with POMDPs
Jungpyo Hong, Jongmin Lee, Kanghoon Lee, Sanggyu Han, Kee-Eung Kim, Il-Chul Moon, and Jae-Hyeon Park
한국정보과학회 학술발표논문집, 2016
Awards and Honors
Outstanding Ph.D. Thesis Award, School of Computing at KAIST, 2022
Qualcomm-KAIST Innovation Awards - Paper Competition, Qualcomm, 2019
Society of Global Ph.D. Fellows Outstanding Presentation Award, 5th SGPF Annual Conference, 2018
Global Ph.D. Fellowship, National Research Foundation of Korea, 2018 ~ 2020
Naver Ph.D. Fellowship, NAVER, 2017
Reviewer
NeurIPS (2016, 2018, 2019, 2020, 2021, 2022)
ICML (2019, 2020,2021, 2022)
AAAI (2020,2021, 2022)
ICLR (2020,2021, 2022)
IJCAI (2021, 2022)
ACML (2017, 2019, 2021)
Machine Learning Journal (2017, 2019)
Journal of Artificial Intelligence Research (2019)
Transactions on Machine Learning Research (2022)
Teaching Experiences
KAIST-Samsung AI Expert Program: Introduction to Reinforcement Learning & Deep Reinforcement Learning TA, KAIST, 2020
KAIST-Samsung AI Expert Program: Introduction to Tensorflow & Reinforcement Learning, TA, KAIST, 2019
Data Structure (CS206), TA, KAIST, 2019
Introduction to Programming (CS101), Head TA, KAIST, 2018
Artificial Intelligence and Machine Learning (CS570), TA, KAIST, Spring 2016
Artificial Intelligence and Machine Learning, TA, KMOOC, Fall 2015
Peer group seminar: Agile web development for non-majors (009.032), Seoul National University, Fall 2012
Extracurricular Activities
Student Representative of School of Computing, KAIST, 2017. 1 ~ 2017. 12
Vice Student Representative of School of Computing, KAIST, 2015. 3 ~ 2016. 2