Wonseok Jeon
Staff Research Scientist at Waymo
San Diego, California, United States
Summary
👤
Senior🎓
Top SchoolWonseok Jeon is a Staff Research Scientist with nine years of experience advancing ML and RL from academia to industry, currently at Waymo after leading ML/RL research at Qualcomm. He holds a PhD in Electrical Engineering from KAIST and built deep expertise in learning-from-demonstration, imitation learning, and inverse RL during postdoctoral work at Mila and McGill. At Qualcomm he delivered practical breakthroughs—applying transformers to DAGs, optimizing peak memory and makespan, and developing speculative decoding for more efficient LLMs. An active contributor to reinforcement learning codebases, he has hands-on experience adapting A3C-style implementations for modern TensorFlow, reflecting both research depth and production engineering rigor. Based in San Diego, he combines multidisciplinary collaboration with a taste for turning theoretical ideas into high-impact systems.
9 years of coding experience
7 years of employment as a software developer
Bachelor of Engineering - BE, Electrical and Electronic Engineering, Bachelor of Engineering - BE, Electrical and Electronic Engineering at Yonsei UniversityDoctor of Philosophy - PhD, Electrical Engineering, Doctor of Philosophy - PhD, Electrical Engineering at Korea Advanced Institute of Science and Technology
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (6)
deep-learning10
tensorflow10
python10
reinforcement-learning10
machine-learning9
rms8
Programming languages (4)
Github contributions (5)
miyosuda/async_deep_reinforce
Jan 2017 - Jan 2017
Asynchronous Methods for Deep Reinforcement Learning
Role in this project:
ML Engineer
ML EngineerContributions:5 commits, 1 PR, 4 comments in 1 day
Contributions summary:Wonseok primarily focused on updating existing code files related to a deep reinforcement learning project. Their commits involved modifying files such as `a3c.py`, `a3c_display.py`, `a3c_training_thread.py`, `game_ac_network.py`, and `rmsprop_applier.py` to align with TensorFlow 0.12. These updates suggest a focus on model training, network architecture, and optimization within the context of asynchronous methods for reinforcement learning. The edits involve code related to network setup, saving and loading models, and training threads, reflecting an active involvement in the project's core functionality.
asynchronousreinforcement-learningasynchronous-methodsdeep-reinforcement-learningreinforcement
wsjeon/SVGD
Aug 2018 - Sep 2018
Contributions:29 commits, 1 push in 17 days
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.