Yue Wu


GHC 8011

4902 Forbes Ave

Pittsburgh, PA 15213

I am a PhD student in the Machine Learning Department at Carnegie Mellon University, currently advised by Dr. Tom Mitchell and Dr. Yuanzhi Li. I also work closely with Dr. Ruslan Salakhutdinov and Dr. Amos Azaria.

I currently work extensively with large language models (LLMs) and reinforcement learning, on embodied reasoning, agent design, and RLHF.

I am generally interested in solving sequential decision problems with information from unstructured natural language (e.g., instruction manuals or news articles).

During my undergraduate at CMU, I have done research in Biorobotics Lab under Dr. Howie Choset and MultiComp Lab under Dr. Louis-Philippe Morency. Recently, I completed my undergraduate thesis on GANs under the supervision of Dr. Zhiting Hu and Dr. Eric Xing. I also worked with Dr. Ruslan Salakhutdinov on sequence generative models in collaboration with Apple Inc.


Feb 28, 2023 I will be joining Microsoft Research over the summer.

latest coverage

selected publications

  1. Read and Reap the Rewards: Learning to Play Atari with the Help of Instruction Manuals
    Yue Wu, Yewen Fan, Paul Pu Liang, and 3 more authors
    In NeurIPS, 2023
    RRL Workshop, ICLR 2022 (oral)
  2. Plan, Eliminate, and Track–Language Models are Good Teachers for Embodied Agents
    Yue Wu, So Yeon Min, Yonatan Bisk, and 5 more authors
    Preprint, 2023
  3. SPRING1.gif
    SPRING: GPT-4 Out-performs RL Algorithms by Studying Papers and Reasoning
    Yue Wu, So Yeon Min, Shrimai Prabhumoye, and 5 more authors
    In NeurIPS, 2023
  4. SmartPlay.png
    SmartPlay : A Benchmark for LLMs as Intelligent Agents
    Yue Wu, Xuan Tang, Tom M. Mitchell, and 1 more author
    Preprint, 2023