← Search

Sensen Gao

2 accepted papers

2025

Adversarial Attacks against Closed-Source MLLMs via Feature Optimal Alignment

NeurIPS 2025poster

Multimodal large language models (MLLMs) remain vulnerable to transferable adversarial examples. While existing methods typically achieve targeted attacks by aligning global features—such as CLIP’s [CLS] token—between adversarial and target samples, they often overlook the rich local information enc…

Cited by 0SourcecodeScholar
2024

Boosting Transferability in Vision-Language Attacks via Diversification along the Intersection Region of Adversarial Trajectory

ECCV 2024poster

"Vision-language pre-training (VLP) models exhibit remarkable capabilities in comprehending both images and text, yet they remain susceptible to multimodal adversarial examples (AEs). Strengthening attacks and uncovering vulnerabilities, especially common issues in VLP models (e.g., high transferabl…