Li-Cheng Lan

UCLA

About me

I am a PhD candidate at UCLA Computer Science department, advised by Prof. Cho-Jui Hsieh. I obtained my Bachelor and Master degree in Computer Science department, National Chiao Tung University (NCTU), Taiwan, supervised by Prof. I-Chen Wu. My research interest is machine learning and agent planning in multi-step environments. I am currently focusing on studying the robustness and the generalization of reinforcement learning (RL), imitation learning (IL), and LLM agents. I am also working on using RL and planning methods for AutoML.

I am looking for a research full-time job in 2025.

Publications (* equal contribution)

Expert Proximity as Surrogate Rewards for Single Demonstration Imitation Learning
Are AlphaZero-like Agents Robust to Adversarial Perturbations?
Learning to schedule learning rate with graph neural networks
Learning to Stop: Dynamic Simulation Monte-Carlo Tree Search
Multiple policy value Monte Carlo tree search
Multilabeled value networks for computer Go

Experiences

  • Applied Scientist Intern, Amazon.com Services LLC (Palo Alto)
  • Member, International Collegiate Programming Contest (ICPC), 44th in 2016 World Finals
  • Member, CGI Go Intelligence (CGI), 2ed place in the 1st World AI Go Open (22,000 USD)

Teaching

TA at UCLA:

  • CS 180 Introduction to Algorithms and Complexity
  • CS 161 Fundamentals of Artificial Intelligence
  • CS 111 Operating System Principles