Paper deep dive
SLAG: A Sensitive Layer Activation-Guided Jailbreak Attack on Vision-Language Models
Jingjing He, Yujie Wang, Yisheng Li, Zhichao Lian
Models: MiniGPT-4
Intelligence
Status: succeeded | Model: google/gemini-3.1-flash-lite-preview | Prompt: intel-v1 | Confidence: 95%
Last extracted: 3/11/2026, 1:13:06 AM
Summary
SLAG is a novel jailbreak attack for Large Vision-Language Models (LVLMs) that utilizes sensitive layer activation guidance to perturb images, achieving a 95% success rate on MiniGPT-4 by targeting specific mid-to-late layers.
Entities (3)
Relation Signals (2)
SLAG → attacks → Large Vision-Language Models
confidence 95% · SLAG, a Sensitive Layer ActivationGuided jailbreak attack for Large Vision-Language Models
SLAG → targets → MiniGPT-4
confidence 95% · On MiniGPT-4, SLAG achieves a 95.0% Attack Success Rate
Cypher Suggestions (0)
No Cypher suggestions yet.
Abstract
We present SLAG, a Sensitive Layer ActivationGuided jailbreak attack for Large Vision-Language Models (LVLMs). SLAG identifies safety-sensitive layers whose activations differ between harmful and benign inputs, and perturbs images using a dual-objective loss to enhance harmful generation while suppressing refusals. On MiniGPT-4, SLAG achieves a 95.0% Attack Success Rate with only 2000 steps, outperforming existing image-only attacks and approaching multimodal methods. Layer analysis shows that a few mid-to-late layers suffice, revealing multiple activation pathways linked to safety failures.
Tags
Links
Full Text
833 characters extracted from source content.
Expand or collapse full text
SLAG: A Sensitive Layer Activation-Guided Jailbreak Attack on Vision-Language Models | IEEE Conference Publication | IEEE Xplore IEEE Account Change Username/Password Update Address Purchase Details Payment Options Order History View Purchased Documents Profile Information Communications Preferences Profession and Education Technical Interests Need Help? US & Canada: +1 800 678 4333 Worldwide: +1 732 981 0060 Contact & Support About IEEE Xplore Contact Us Help Accessibility Terms of Use Nondiscrimination Policy Sitemap Privacy & Opting Out of Cookies A not-for-profit organization, IEEE is the world's largest technical professional organization dedicated to advancing technology for the benefit of humanity.© Copyright 2026 IEEE - All rights reserved. Use of this web site signifies your agreement to the terms and conditions.