Tingwei Zhang
Department of Computer Science Cornell University
Welcome! I am a Ph.D. candidate in Computer Science at Cornell University, advised by Prof. Vitaly Shmatikov. My research focuses on the security and privacy aspects of machine learning. Previously, I graduated from University of Virginia with a B.A. in Computer Science, where I was very fortunate to work with Prof. Yuan Tian and Prof. David Evans.
Current News:
-📌 Gave a talk titled “How to Steal Reasoning Without Reasoning Traces” at the UVA Artificial Intelligence and Machine Learning (AIML) seminar.
-📌 Gave a talk titled “Adversarial Examples in Multi-modal Systems” in CS 684: Trustworthy and Responsible AI at the University of Massachusetts.
Publications
(2026).
Speculative Probing: LLM Monitoring at Speculative-Decoding Cost.
Preprint, 2026.
(2026).
Deep-Research Agents Can Be Poisoned via User-Generated Content.
Preprint, 2026.
PDF Cite Code Tweet 404 Media Nature Le Monde NRC El PaÃs Corriere della Sera
(2026).
Adversarial Hubness in Multi-Modal Retrieval.
S&P 2026.
(2026).
How to Steal Reasoning Without Reasoning Traces.
Preprint, 2026.
(2026).
Adversarial Decoding: Generating Readable Documents for Adversarial Objectives.
Findings of EACL 2026.
(2025).
Self-interpreting Adversarial Images.
USENIX Security, 2025.
(2024).
Adversarial Illusions in Multi-Modal Embeddings.
USENIX Security, 2024 (Distinguished Paper Award).