About

Kimin Lee is a prominent researcher specializing in deep reinforcement learning (RL), with a particular focus on generalization, model-based RL, and human-in-the-loop learning systems. His work addresses some of the most pressing challenges in deploying RL agents in real-world settings, where environments are dynamic, data is scarce, and reward engineering is costly. Lee's most recognized contribution, "Network Randomization: A Simple Technique for Generalization in Deep Reinforcement Learning," has accumulated nearly 100 citations across its versions, establishing him as a leading voice in improving RL robustness to unseen environments. He further extended this theme through information bottleneck approaches and context-aware dynamics models, tackling the critical problem of dynamics generalization across varying conditions. His research portfolio also spans model-based RL, one-shot visual imitation learning, and preference-based RL — where his PEBBLE and SURF frameworks demonstrate innovative strategies for reducing human feedback requirements while maintaining training efficiency. His offline-to-online RL work bridges the gap between static datasets and adaptive real-world deployment. With over 230 cumulative citations and contributions spanning robotics, sample efficiency, and human-AI interaction, Lee's research offers foundational tools for building more adaptable and practically deployable intelligent agents.

Research Focus

Key Achievements

9
H-Index
18
Papers
272
Total Citations
15
Avg Citations/Paper
🏆 Most Cited Paper
Network Randomization: A Simple Technique for Generalization in Deep Reinforcement Learning
48 citations · 2020
📈 Most Prolific Year: 2022 (5 Papers)
🤝 Key Collaborators: 36
🏛 Institutions: University of California, Berkeley, Berkeley College, Korea Advanced Institute of Science and Technology, Korea Innotech (South Korea)

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 14 days ago