cv

Basics

Name Hongxiang Zhang
Label Research assistant
Email hxxzhang@ucdavis.edu
Phone (530) 564-2580
Url https://harrison4ride.github.io

Work

  • Jan.2023 - Present
    Teaching Assistant
    UC Davis
    • ECS 140 Promgramming Langugae.
    • ECS 36C Data Structure, algorithms and analytics.
    • ECS 153 Computer Security.
  • Feb.2022 - Jul.2022
    Software and Operation Engineer
    Intern, Volkswagen Group China
    • Achieved data masking and auto-populated to JIRA log in Python, and reduced the workload of the IT service team, by over 30%.
    • Intuitively analyzed and provided services and finance data to CIO with Tableau, helping the management level adjust the company service strategy agilely.
    • Supported team, organized weekly meetings to track team's progress, promptly communicated with vendors as IT interfaces and helped solve over 30 emergency cases.
  • Dec.2020 - Feb.2021
    Quality Assurance
    Intern, Didi
    • Developed distributed Java invoice services to enable the interaction between server and end-user devices that served more than 1 million users per day.
    • Achieved unit testing automation by using JUnit and the test case in Redis, reduced over 40% of the testing engineer's workload.

Education

  • Sep.2022 - Present

    Davis, California

    Master
    University of California, Davis
    Computer Science
  • Sep.2018 - Jun.2022

    Shandong, China

    Bachelor
    Shandong University
    Computer Science
  • Aug.2020 - Jul.2022

    Canberra, Australia

    Bachelor (Honours)
    Australian National University
    Computer Science

Certificates

Azure Fundamentals
Microsoft Apr-2022

Publications

  • Oct.2024
    SteerDiff: Steering towards Safe Text-To-Image Diffusion Model
    SteerDiff is introduced, a lightweight adaptor module designed to act as an intermediary between user input and the diffusion model, ensuring that generated images adhere to ethical and safety standards with little to no impact on usability.
  • Jun.2024
    LLAMAFUZZ: Large Language Model Enhanced Greybox Fuzzing
    This paper uses the pre-trained knowledge of LLM about data conversion and format to generate new valid inputs to enhance greybox fuzzing for structured data and fine-tuned it with paired mutation seeds to learn structured format and mutation strategies effectively.