Juzheng Zhang

2nd-year PhD student in Computer Science

Juzheng Zhang's Profile Image

I am a 2nd-year PhD student in Computer Science at the University of Maryland, College Park, advised by Prof. Tom Goldstein. My research focuses on post-training methods for LLMs and agentic systems, with an emphasis on RL-based alignment and continual learning to improve reasoning reliability and enable stable self-improvement over time. Before joining UMD, I graduated with honors from Shanghai Jiao Tong University, where I earned a Bachelor’s degree in Information Engineering.

I am currently seeking research internship opportunities for summer 2026. Feel free to reach out! 🤝

Selected Works

  1. Blog
    logit-fusion.png
    Learning from Mixed Rollouts: Logit Fusion as a Bridge Between Imitation and Exploration
    Juzheng Zhang, Abhimanyu Hans, John Kirchenbauer, Micah Goldblum, Ashwinee Panda, and 1 more author
    Notion Blog, 2026
  2. COLM
    lori.png
    LoRI: Reducing Cross-Task Interference in Multi-Task Low-Rank Adaptation
    Juzheng Zhang, Jiacheng You, Ashwinee Panda, and Tom Goldstein
    Conference on Language Modeling, 2025
  3. arXiv
    unimot.png
    UniMoT: Unified Molecule-Text Language Model with Discrete Token Representation
    Juzheng Zhang, Yatao Bian, Yongqiang Chen, and Quanming Yao
    arXiv preprint arXiv:2408.00863, 2024
  4. KDD
    hl-gnn.png
    Heuristic Learning with Graph Neural Networks: A Unified Framework for Link Prediction
    Juzheng Zhang, Lanning Wei, Zhen Xu, and Quanming Yao
    In Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 2024

Honors and Awards

  • Dean’s Fellowship, University of Maryland, 2024
  • Outstanding Graduate, Shanghai Jiao Tong University, 2023
  • Shanghai Government Scholarship, Shanghai Government, 2021
  • Meritorious Winner, International Mathematical Contest in Modeling, 2021

Services

  • Reviewer for NeurIPS 25, ICML 25 & 24, ICLR 26 & 25 & 24, KDD 24, and AAAI 26.
  • Teaching Assistant for CMSC 250 Discrete Structures, Fall 2024