I am working as a Staff Machine Learning Engineer at Qualcomm AI Research. Previously I was a postdoctoral researcher at Mila, McGill University under the supervision of Prof. Joelle Pineau and at KAIST under the supervision of Prof. Kee-Eung Kim. I have been working on learning to control, mostly on learning from demonstrations, e.g., imitation learning, inverse reinforcement learning, and offline reinforcement learning.

For more information, please check out my CV and research statement.

Contact Information: jeonwonseok0125 [at] gmail [dot] com


Publications

[P1] Wonseok Jeon, Paul Barde, Derek Nowrouzezahrai, Joelle Pineau,
“Scalable multi-agent inverse reinforcement learning via actor-attention-critic,”
preprint, 2020 [paper]

[C9, W5] Jongmin Lee*, Wonseok Jeon*, Byung-Jun Lee, Joelle Pineau, Kee-Eung Kim,
“OptiDICE: Offline Policy Optimization via Stationary Distribution Correction Estimation,”
ICML, 2021
ICLR Workshop on A Roadmap to Never-Ending RL (NERL), 2021

[C8, W4] Wonseok Jeon, Chen-Yang Su, Paul Barde, Thang Doan, Derek Nowrouzezahrai, Joelle Pineau,
“Regularized inverse reinforcement learning,”
ICLR, 2021 Spotlight Presentation, 167/2997=5.57%
NeurIPS Deep Reinforcement Learning Workshop (DRLW), 2020
[paper]

[C7] Paul Barde*, Julien Roy*, Wonseok Jeon*, Joelle Pineau, Christopher Pal, Derek Nowrouzezahrai,
“Adversarial soft advantage fitting: Imitation learning without policy optimization,”
NeurIPS, 2020 Spotlight Presentation, 395/9454=4.07%
(*Equal Contribution)
[paper]

[W3] Wonseok Jeon, Paul Barde, Derek Nowrouzezahrai, and Joelle Pineau,
“Scalable and sample-efficient multi-agent imitation learning,”
AAAI Workshop on Reinforcement Learning in Games, 2020
[paper]

[C6] Jongmin Lee, Wonseok Jeon, Geon-Hyeong Kim, and Kee-Eung Kim,
“Monte-Carlo tree search in continuous action spaces with value gradients,”
AAAI, 2020
[paper]

[C5] Geon-Hyeong Kim, Youngsoo Jang, Jongmin Lee, Wonseok Jeon, Hongseok Yang, and Kee-Eung Kim,
“Trust region sequential variational inference,”
ACML, 2019
[paper]

[C4] Wonseok Jeon, Seokin Seo, and Kee-Eung Kim,
“A Bayesian approach to generative adversarial imitation learning,”
NeurIPS, 2018 Spotlight Presentation, 168/4856=3.46%
[paper], [code]

Publications - Information Theory and Wireless Communication

[J3] Wonseok Jeon and Sae-Young Chung,
“Capacity of continuous-space electromagnetic channels with lossy transceiver,”
IEEE Transactions on Information Theory, Mar. 2018
[paper]

[J2] Jin Hak Kim, Wonseok Jeon and Sae-Young Chung,
“Asymptotic Analysis on Directivity and Beamwidth of Uniform Circular Array,”
IEEE Antennas and Wireless Propagation Letters, Oct. 2017
[paper]

[J1] Wonseok Jeon, Jin Hak Kim and Sae-Young Chung,
“Effect of mutual coupling on uniform circular arrays with vector antenna elements,”
IEEE Antennas and Wireless Propagation Letters, Feb. 2017
[paper]

[W2] Wonseok Jeon and Sae-Young Chung,
“Interference mitigation using receiver superdirectivity,”
ITA Workshop, 2016

[C3] Wonseok Jeon and Sae-Young Chung,
“Interference mitigation utilizing antenna mutual coupling,”
Asimolar Conference on Signals, Systems and Computers, 2015
[paper]

[C2] Wonseok Jeon and Sae-Young Chung,
“Improving degrees of freedom of wireless channels using superdirectivity,”
ISIT, 2015
[paper]

[W1] Wonseok Jeon and Sae-Young Chung,
“Noise spatial correlation and receive superdirectivity in wireless channels,”
ITA Workshop, 2015

[C1] Wonseok Jeon and Sae-Young Chung,
“The capacity of wireless channels: A physical approach,”
ISIT, 2013
[paper]


Reviewer

Neural Information Processing Systems (NeurIPS)
2019 / 2020

AAAI Conference on Artificial Intelligence (AAAI)
2019 / 2021

International Conference of Learning Representations (ICLR)
2020 / 2021

Asian Conference on Machine Learning (ACML)
2019 / 2020


Talks

“A Bayesian approach to generative adversarial imitation learning,”
SK T-Brain
Mar, 2019


Open Source Activity

“Multi-Agent Deterministic Deep Policy Gradient (MADDPG)”
Ray/RLlib,
[code1], [code2]


Awards

“Superdirectivity in Wireless Channels”
Qualcomm-KAIST Innovation Award, 2015 [QCI Award-2015-07]