Experience

  1. Summer Intern, AI Team

    The Washington Post

    2026 Engineering & Business Operations Internship Program, Washington, D.C.

    • Developed an LLM-assisted representation framework for personalized news recommendation, introducing StoryID to capture the underlying story, entities, and article intent beyond article-level IDs or coarse topics.
    • Integrated StoryID representations with existing recommendation models to improve personalization, with evaluation focused on cold-start recommendation, content diversity, and recommendation interpretability.

  2. Graduate Research Assistant

    George Mason University

    Cyber Security Engineering, advised by Dr. Zhuangdi Zhu.

    • Research on LLM post-training, safety alignment, machine unlearning, and agentic AI.
    • First-author work published at ICLR 2026 and EMNLP 2025 Main Conference; co-authored work at EMNLP 2026 Main Conference.
    • Reviewer for ICLR 2025/2026 and IEEE Transactions on Information Forensics and Security (TIFS).

  3. Research Assistant

    State Key Laboratory of Information Security, Chinese Academy of Sciences
    • Developed PROFL, a privacy-preserving federated learning framework with stringent defense against poisoning attacks (published at CSCWD 2024).
    • Combined two-trapdoor homomorphic encryption with statistics-based outlier filtering to defend against concealed impersonation attacks.

Education

  1. PhD Information Technology (Cybersecurity)

    George Mason University

    GPA: 4.00/4.00 — expected to graduate in 2028.

    Research on LLM post-training, safety alignment, machine unlearning, and agentic AI. Advised by Dr. Zhuangdi Zhu.

  2. MSc Cyber Security

    University of Chinese Academy of Sciences

    GPA: 3.56/4.00

    Coursework in AI security, security protocols, cryptography, and privacy-preserving machine learning. My thesis work sat at the intersection of machine learning and information security: defending federated learning against poisoning attacks while preserving user privacy, published at CSCWD 2024.

  3. BSc Computer Science

    Harbin University of Science and Technology

    GPA: 3.84/4.00 (Top 1%)

    Foundational coursework in computer science and machine learning. Graduated in the top 1% of my cohort and earned direct admission to a Master’s program at the Chinese Academy of Sciences.

Skills & Hobbies
Technical Skills
Python
Data Science
LLM Post-Training & Alignment
PyTorch / Hugging Face
Hobbies
Tennis / American Football
Workout
Photography
Awards
Meritorious Winner (First Prize), Mathematical Contest in Modeling (MCM) 2020
Consortium for Mathematics and Its Applications (COMAP) ∙ April 2020
Led a team to develop a 3D cellular automaton model and optimize it using genetic algorithms to solve an open-ended problem. Organized regular pre-competition meetings and simulated contests to enhance team collaboration. Took charge of developing mathematical models, algorithm design, and implementation. Recognized for exceptional teamwork, leadership, and technical contributions.