🧐 About Me

Hi there! I am a 3-year PhD student in Computer Science at the ETH Zurich, under the supervision of Prof. Florian Tramèr, and a member of the Secure and Private AI (SPY) Lab.

Research Interests

🔍 I'm drawn to problems where...
🤔 Something could go wrong with LLMs
📊 We can perform rigorous evaluation
⚔️ We can provide a stronger attack
🎯 It's a realistic threat
LLM Safety and Security Prompt Injection LLM Optimization LLM Alignment Adversarial Example Privacy Evaluation Membership Inference Attacks Synthetic Data

🔥 News

2025.09 🎉 RealMath is accepted by NeurIPS 2025
2024.09 🎉 AgentDojo is accepted by NeurIPS 2024 (dataset and benchmark track). Benchmark
2024.01 🎉 Real-Fake is accepted by ICLR 2024
2023.03 🎉 I graduate from ZJU

📒 Blogs

Our lab has very nice 📚 Blogs about AI security and privacy, highly recommended for reading!

📝 Selected Publications

( * indicates equal contribution. Full list of publications)

📚 Preprint
TBD

🚀 Something is Coming Soon™ (Probably) Status: Thinking hard 🤔 …]

preprint
sym

Black-box Optimization of LLM Outputs by Asking for Directions

Jie Zhang, Meng Ding, Yang Liu, Jue Hong, Florian Tramèr

code

IEEE SP 2025, DLSP workshop

Position: Adversarial ML Problems Are Getting Harder to Solve and to Evaluate

Javier Rando*, Jie Zhang*, Nicholas Carlini, Florian Tramèr

[IEEE SP 2025, DLSP workshop]

✅ Accepted
NeurIPS 2025
sym

RealMath: A Continuous Benchmark for Evaluating Language Models on Research-Level Mathematics

Jie Zhang, Cezara Petrui, Kristina Nikolić, Florian Tramèr

[NeurIPS 2025, Dataset $\&$ Benchmark Track]

ICML 2025
sym

The Jailbreak Tax: How Useful are Your Jailbreak Outputs?

Kristina Nikolić, Luze Sun, Jie Zhang, Florian Tramèr

[ICML 2025, spotlight]

SaTML 2025
sym

Position: Membership Inference Attacks Cannot Prove that a Model Was Trained On Your Data

Jie Zhang, Debeshee Das, Gautam Kamath, Florian Tramèr

[IEEE SaTML 2025]

CCS 2024
sym

Evaluations of Machine Learning Privacy Defenses are Misleading

Michael Aerni*, Jie Zhang*, Florian Tramèr

[ACM CCS 2024]

ICLR 2025
sym
IEEE SP 2025, DLSP workshop
sym

Blind Baselines Beat Membership Inference Attacks for Foundation Models

Debeshee Das, Jie Zhang, Florian Tramèr

[IEEE SP 2025, DLSP workshop]

NeurIPS 2024
sym

AgentDojo: A Dynamic Environment to Evaluate Prompt Injection Attacks and Defenses for LLM Agents

Edoardo Debenedetti, Jie Zhang, Mislav Balunović, Luca Beurer-Kellner, Marc Fischer, Florian Tramèr

[NeurIPS 2024 Dataset $\&$ Benchmark Track]

📖 Education

🎓 PhD
Computer Science
🇨🇭 Switzerland
Ongoing
🎯 MSc
Software Engineering
🇨🇳 China
Mar 2023
Advisor: Prof. Chao Wu
🏆 BSc
Internet of Things
🇨🇳 China
Jul 2020
Bachelor's Degree

🎤 Talks

  • ResearchTrend Connect (2024.12)
    "Membership Inference Attacks Cannot Prove that a Model Was Trained On Your Data" [paper]
  • Google, Differential Privacy for ML (2025.04)
    "Membership Inference Attacks Cannot Prove that a Model Was Trained On Your Data" [paper]

🎖 Honors and Awards

  • 2021.05 We won the first prize on CVPR21 Workshop (Adversarial Machine Learning in Real-World Computer Vision Systems and Online Challenges, rank: 1 / 1558).
  • 2022.10 China National Scholarship, Zhejiang University, 2022
  • Outstanding Student Scholarship, First Prize, Hainan University, 2018, 2019, 2020.

💬 Services

  • Journal Reviewer:
    • IEEE Transactions on Neural Networks and Learning Systems
    • Neural Networks
    • IEEE Transactions on Pattern Analysis and Machine Intelligence
  • Conference Reviewer: ICLR, AAAI, CVPR, ICML, ECCV, ICCV, NeurIPS.

💻 Internships

🎙 Miscellaneous

Travel

I enjoy the time traveling with my families and friends. I am always excited about visiting new places and knowing different cultures.

My cat

My wife and I have three cats together, they are very adorable and have brought a lot of fun to our lives!

图片名称 图片名称 图片名称