Qinyuan Wu

qinyuan.JPG

qwu [at] mpi-sws [dot] org

Campus E1 5

66125, Saarbruecken, Germany

I am a second-year PhD student at CS@Max Planck and Max Planck Institute for Software System (MPI-SWS), under the guidance of Prof. Krishna Gummadi. I am also fortunate to closely collaborate with and be advised by Prof. Evimaria Terzi in Boston University. My research focuses on understanding the inner workings of deep learning models and establishing trustworthy AI systems. Currently, I am investigating methods for estimating and understanding latent knowledge in Large Language Models (LLMs).

Prior to joining MPI-SWS, I earned my bachelor’s degree in mathematics-physics fundamental science from the Yingcai Honors College at the University of Electronic Science and Technology of China (UESTC).

news

Oct 10, 2022 I joined MPI-SWS :slightly_smiling_face:

selected publications

  1. Towards Reliable Latent Knowledge Estimation in LLMs: In-Context Learning vs. Prompting Based Factual Knowledge Extraction
    Qinyuan Wu, Mohammad Aflah Khan, Soumi Das, and 7 more authors
    arXiv preprint arXiv:2404.12957, 2024
  2. Understanding Memorisation in LLMs: Dynamics, Influencing Factors, and Implications
    Till Speicher, Mohammad Aflah Khan, Qinyuan Wu, and 5 more authors
    arXiv preprint arXiv:2407.19262, 2024