DAVSP: Safety Alignment for Large Vision-Language Models via Deep Aligned Visual Safety Prompt

Yitong Zhang; Jia Li; Liyi Cai; Ge Li

AAAI 2026technical0 citations

DAVSP: Safety Alignment for Large Vision-Language Models via Deep Aligned Visual Safety Prompt

Abstract

Large Vision-Language Models (LVLMs) have achieved impressive progress across various applications but remain vulnerable to malicious queries. Existing safety alignment approaches typically fail to resist malicious queries while preserving utility on benign ones effectively. To address these challenges, we propose DAVSP, which is built upon two key innovations. First, we introduce Visual Safety Prompt, which appends a trainable padding region around the input image. It preserves visual features and expands the optimization space. Second, we propose Deep Alignment, a novel approach to train the visual safety prompt through supervision in the model

BibTeX

@inproceedings{aaai2026_davspsafetyalign,
  title = {DAVSP: Safety Alignment for Large Vision-Language Models via Deep Aligned Visual Safety Prompt},
  author = {Yitong Zhang and Jia Li and Liyi Cai and Ge Li},
  booktitle = {AAAI 2026},
  year = {2026}
}