2 papers
cs.CR2026
TIER: Trajectory-Invariant Explanation Regularization for Membership Privacy
Varun Sharma, Kar Wai Fok, Vrizlynn L. L. Thing
Explainability is central to building trustworthy AI, yet explanation interfaces can inadvertently provide adversaries with an expanded privacy-related attack surfaces. Recent stud…
cs.CR2026
DE-FIVE: Detecting Malicious Image Prompts via Fourier Features and Image Vector Embeddings
Xingwei Zhong, Varun Sharma, Kar Wai Fok +1
Vision language models (VLMs) employ both visual and textual modalities to enable advanced vision-language inference. However, incorporating visual modalities expands the attack su…