Haonan Shi
503 Olin, 2001–2017 Martin Luther King Jr Dr
Cleveland, OH 44106
United States
I am a third-year Ph.D. candidate at Case Western Reserve University, advised by Prof. An Wang and co-advised by Dr. Tu Ouyang. My research focuses on Large Language Models Safety and Machine Learning Privacy.
Prior to this, I received my bachelor’s degree from South China University of Technology, where I conducted research on privacy under the supervision of Prof. Hongyun Xu.
I am passionate about applying my research to real-world problems and look forward to internship opportunities and research collaborations.
Email: haonan.shi3[AT]case.edu(please replace “[AT]” by “@”)
research interests
My research lies at the intersection of Machine Learning Safety(Security) and Privacy. Specifically, the focus areas include:
Privacy-Preserving Machine Learning: Developing effective and efficient methods to detect privacy vulnerabilities in machine learning models and systems through membership inference attacks [EuroS&P, PoPETs]. Also investigating privacy-preserving fine-tuning approaches for large language models [WWW].
Large Language Model Safety: Exploring reasoning-enhanced safety alignment methods that effectively detect and defend against jailbreak attacks while preserving both inference-time efficiency and model performance [Under Review].
news
| May 21, 2025 | Our paper titled Unveiling Client Privacy Leakage from Public Dataset Usage in Federated Distillation was accepted by PoPETs 2025. |
|---|---|
| Mar 05, 2025 | Our paper titled Navigating the Designs of Privacy-Preserving Fine-tuning for Large Language Models will also be a presentation at the ICLR25 FM-Wild workshop. |
| Jan 20, 2025 | Our paper titled Navigating the Designs of Privacy-Preserving Fine-tuning for Large Language Models was accepted by WWW 2025. |
| Mar 26, 2024 | Our paper titled Learning-Based Difficulty Calibration for Enhanced Membership Inference Attacks was accepted by EuroS&P 2024. |
