Scholar

Siyuan Cheng

Google Scholar ID: GcL9AFMAAAAJ

Purdue University

AI SecurityAdversarial Machine Learning

Citations & Impact

All-time

Citations

1,012

H-index

i10-index

Publications

Co-authors

list available

Contact

Publications

20 items

Browse publications on Google Scholar (top-right) ↗

Resume (English only)

Academic Achievements

Published 'BAIT: Large Language Model Backdoor Scanning by Inverting Attack Target' at IEEE S&P 2025 (co-first author).
Published 'BiScope: AI-generated Text Detection by Checking Memorization of Preceding Tokens' at NeurIPS 2024.
Published 'CENSOR: Defense Against Gradient Inversion via Orthogonal Subspace Bayesian Sampling' at NDSS 2025.
Multiple papers at IEEE S&P 2024, including 'ODSCAN' (co-first author) and 'Exploring the Orthogonality and Linearity of Backdoor Attacks' (co-first author).
Published 'Rethinking the Invisible Protection against Unauthorized Image Usage in Stable Diffusion' at USENIX Security 2024.
Authored preprints such as 'Rapid Optimization for Jailbreaking LLMs via Subconscious Exploitation and Echopraxia' (co-first author).

Co-authors

12 total