thumbnail image

Kevin Li

  • Home
  • CS 189
  • Hello!

    I'm a master's student at UC Berkeley, advised by Professor Sergey Levine in the Robotic AI & Learning Lab (RAIL). Previously, I also did my undergrad in EECS at Berkeley.

     

    I'm excited about machine learning, education, history, photography, spicy food, exploring new places, and taking naps.

  • Teaching

    I love teaching and have been actively involved in the education community at Berkeley for a while! Here are some classes I've been involved with:

    • Spring 2022: CS 189/289A, Machine Learning (Head TA)
    • Fall 2021: CS 285, Deep Reinforcement Learning (TA)
    • Spring 2021: CS 189/289A, Machine Learning (Head TA)
    • Spring 2020: CS 189/289A, Machine Learning (20-hr TA)
    • Spring 2019: CS 61A, Intro to Programming (TA)
    • Fall 2018: CS 61A, Intro to Programming (TA)
    • Summer 2018: CS 61A, Intro to Programming (Head of Projects)

    Work Experience

    • (Post-graduation: Software Engineering @ Applied Intuition)
    • Summer 2021: Perception/ML Infra @ Waymo
    • Summer 2020: ML research @ Robotic AI & Learning Lab (RAIL)
    • Summer 2019: Backend/payments infra @ Stripe

    Research

    I'm interested in algorithms to make reinforcement learning feasible in real-world domains such as robotics, autonomous vehicles, and medicine. My current focus is on developing better techniques for uncertainty estimation on deep neural networks, which can be used to improve exploration, goal-driven RL, and offline RL.

     

    Publications

    MURAL: Meta-Learning Uncertainty-Aware Rewards for Outcome-Driven Reinforcement Learning
    Kevin Li*, Abhishek Gupta*, Ashwin D Reddy, Vitchyr Pong, Aurick Zhou, Justin Yu, Sergey Levine
    International Conference on Machine Learning (ICML) 2021

    Paper | Website | Github

     

    Simulating Polyculture Farming to Tune Automation Policies for Plant Diversity and Precision Irrigation

    Yahav Avigal, Jensen Gao, William Wong, Kevin Li, Grady Pierroz, Fang Shuo Deng, Mark Theis, Mark Presten, Ken Goldberg

    IEEE Conference on Automation Science and Engineering (CASE) 2020 - Best Student Paper Award

    Paper | Github

     

    Cookie Use
    We use cookies to ensure a smooth browsing experience. By continuing we assume you accept the use of cookies.
    Learn More