Junyi Zhu (朱俊屹)

Senior researcher at Samsung Electronics R&D Institute UK (SRUK).

job_photo_selected.png

I’m currently working at SRUK with a focus on advanced AI algorithm research. I’ve done my PhD studies at KU Leuven in Belgium, under the guidance of Prof. Matthew Blaschko. With great passion and curiosty for AI technologies, I have contribued to many domains, including image generation models, large language models, distributed learning and privacy-preserving machine learning.

Before my PhD, I earned my Master’s degree from the Karlsruhe Institute of Technology in Germany, specializing in information technologies and vehicle engineering. During my Master’s program, I worked on several research projects using AI to solve control and perception tasks of autonomous driving at Institute of Measurement and Control Systems and Research Center for Information Technology in Karlsruhe.

Engaging in AI research brings me immense joy, and I strive to utilize this technology to enhance industry or our daily lives. I look forward to the day when AGI becomes a reality.

If you are interested in research collaboration, feel free to reach out!😊

news

Mar 14, 2025 Our paper Linear Combination of Saved Checkpoints Makes Consistency and Diffusion Models Better has been accepted at ICLR 2025.
Nov 27, 2024 Thrilled to announce that I have successfully earned my PhD in AI from KU Leuven—grateful for the journey, mentorship, and support along the way!
Sep 16, 2024 I’m exited to be joining Samsung Research UK!

selected publications

  1. ICLR
    lcsc.png
    Linear Combination of Saved Checkpoints Makes Consistency and Diffusion Models Better
    Enshu Liu*, Junyi Zhu*, Zinan Lin, and 7 more authors
    2025
    * = Co-first authors
    HTML
  2. EMNLP Findings
    fastmem.png
    FastMem: Fast Memorization of Prompt Improves Context Awareness of Large Language Models
    Junyi Zhu*, Shuochen Liu*, Yu Yu, and 6 more authors
    2024
    * = Co-first authors
    HTML
  3. CVPR
    pfedvem.png
    Confidence-aware Personalized Federated Learning via Variational Expectation Maximization
    Junyi Zhu*, Xingchen Ma*, and Matthew B. Blaschko
    2023
    * = Co-first authors
    HTML
  4. ICML
    sme.png
    Surrogate Model Extension (SME): A Fast and Accurate Weight Update Attack on Federated Learning
    Junyi Zhu, Ruicong Yao, and Matthew B. Blaschko
    2023
    HTML
  5. TMLR
    nerf.png
    Implicit Neural Representations for Robust Joint Sparse-View CT Reconstruction
    Jiayang Shi*, Junyi Zhu*, Daniel M. Pelt, and 2 more authors
    2024
    * = Co-first authors
    HTML