Copertina

Bridging the Gap in Vision Language Models in Identifying Unsafe Concepts Across Modalities

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Qu, Yiting, Backes, Michael, Zhang, Yang
Natura:	Preprint
Pubblicazione:	2025
Soggetti:	Cryptography and Security Artificial Intelligence
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Sparse Models, Sparse Safety: Unsafe Routes in Mixture-of-Experts LLMs
di: Jiang, Yukun, et al.
Pubblicazione: (2026)

Understanding LLM Behavior When Encountering User-Supplied Harmful Content in Harmless Tasks
di: Chu, Junjie, et al.
Pubblicazione: (2026)

Concept-Guided Backdoor Attack on Vision Language Models
di: Shen, Haoyu, et al.
Pubblicazione: (2025)

Probing the Safety Response Boundary of Large Language Models via Unsafe Decoding Path Generation
di: Wang, Haoyu, et al.
Pubblicazione: (2024)

HarmfulSkillBench: How Do Harmful Skills Weaponize Your Agents?
di: Jiang, Yukun, et al.
Pubblicazione: (2026)

HomeSafe-Bench: Evaluating Vision-Language Models on Unsafe Action Detection for Embodied Agents in Household Scenarios
di: Pu, Jiayue, et al.
Pubblicazione: (2026)

AttackPilot: Autonomous Inference Attacks Against ML Services With LLM-Based Agents
di: Wu, Yixin, et al.
Pubblicazione: (2025)

Bridging the Copyright Gap: Do Large Vision-Language Models Recognize and Respect Copyrighted Content?
di: Xu, Naen, et al.
Pubblicazione: (2025)

SAGE: Exploring the Boundaries of Unsafe Concept Domain with Semantic-Augment Erasing
di: Zhu, Hongguang, et al.
Pubblicazione: (2025)

NEXUS: Network Exploration for eXploiting Unsafe Sequences in Multi-Turn LLM Jailbreaks
di: Asl, Javad Rafiei, et al.
Pubblicazione: (2025)

On the Proactive Generation of Unsafe Images From Text-To-Image Models Using Benign Prompts
di: Wu, Yixin, et al.
Pubblicazione: (2023)

Membership Inference Attacks Against Vision-Language Models
di: Hu, Yuke, et al.
Pubblicazione: (2025)

Prompt Stealing Attacks Against Text-to-Image Generation Models
di: Shen, Xinyue, et al.
Pubblicazione: (2023)

UnsafeBench: Benchmarking Image Safety Classifiers on Real-World and AI-Generated Images
di: Qu, Yiting, et al.
Pubblicazione: (2024)

Less Is More -- Until It Breaks: Security Pitfalls of Vision Token Compression in Large Vision-Language Models
di: Zhang, Xiaomei, et al.
Pubblicazione: (2026)

Real Money, Fake Models: Deceptive Model Claims in Shadow APIs
di: Zhang, Yage, et al.
Pubblicazione: (2026)

The Security Threat of Compressed Projectors in Large Vision-Language Models
di: Zhang, Yudong, et al.
Pubblicazione: (2025)

Seeing is Deceiving: Exploitation of Visual Pathways in Multi-Modal Language Models
di: Janowczyk, Pete, et al.
Pubblicazione: (2024)

Multi-turn Jailbreaking Attack in Multi-Modal Large Language Models
di: Das, Badhan Chandra, et al.
Pubblicazione: (2026)

Do Vision-Language Models Respect Contextual Integrity in Location Disclosure?
di: Yang, Ruixin, et al.
Pubblicazione: (2026)

JADES: A Universal Framework for Jailbreak Assessment via Decompositional Scoring
di: Chu, Junjie, et al.
Pubblicazione: (2025)

Adversarial attacks against Modern Vision-Language Models
di: La Torre, Alejandro Paredes
Pubblicazione: (2026)

Latent Fusion Jailbreak: Blending Harmful and Harmless Representations to Elicit Unsafe LLM Outputs
di: Xing, Wenpeng, et al.
Pubblicazione: (2025)

Reconstruct Your Previous Conversations! Comprehensively Investigating Privacy Leakage Risks in Conversations with GPT Models
di: Chu, Junjie, et al.
Pubblicazione: (2024)

Multi-PA: A Multi-perspective Benchmark on Privacy Assessment for Large Vision-Language Models
di: Zhang, Jie, et al.
Pubblicazione: (2024)

Towards Understanding Unsafe Video Generation
di: Pang, Yan, et al.
Pubblicazione: (2024)

Breach By A Thousand Leaks: Unsafe Information Leakage in `Safe' AI Responses
di: Glukhov, David, et al.
Pubblicazione: (2024)

$\texttt{ModSCAN}$: Measuring Stereotypical Bias in Large Vision-Language Models from Vision and Language Modalities
di: Jiang, Yukun, et al.
Pubblicazione: (2024)

JailDAM: Jailbreak Detection with Adaptive Memory for Vision-Language Model
di: Nian, Yi, et al.
Pubblicazione: (2025)

Automated Post-Incident Policy Gap Analysis via Threat-Informed Evidence Mapping using Large Language Models
di: Oh, Huan Lin, et al.
Pubblicazione: (2026)

When Benign Inputs Lead to Severe Harms: Eliciting Unsafe Unintended Behaviors of Computer-Use Agents
di: Jones, Jaylen, et al.
Pubblicazione: (2026)

MixBridge: Heterogeneous Image-to-Image Backdoor Attack through Mixture of Schrödinger Bridges
di: Qin, Shixi, et al.
Pubblicazione: (2025)

"Humans welcome to observe": A First Look at the Agent Social Network Moltbook
di: Jiang, Yukun, et al.
Pubblicazione: (2026)

PromptGuard: Soft Prompt-Guided Unsafe Content Moderation for Text-to-Image Models
di: Yuan, Lingzhi, et al.
Pubblicazione: (2025)

Benchmark of Benchmarks: Unpacking Influence and Code Repository Quality in LLM Safety Benchmarks
di: Chu, Junjie, et al.
Pubblicazione: (2026)

On the Feasibility of Using MultiModal LLMs to Execute AR Social Engineering Attacks
di: Bi, Ting, et al.
Pubblicazione: (2025)

Unsafe LLM-Based Search: Quantitative Analysis and Mitigation of Safety Risks in AI Web Search
di: Luo, Zeren, et al.
Pubblicazione: (2025)

Distilling Lightweight Language Models for C/C++ Vulnerabilities
di: Wei, Zhiyuan, et al.
Pubblicazione: (2025)

Large Language Models for Power System Security: A Novel Multi-Modal Approach for Anomaly Detection in Energy Management Systems
di: Zaboli, Aydin, et al.
Pubblicazione: (2025)

Visual-RolePlay: Universal Jailbreak Attack on MultiModal Large Language Models via Role-playing Image Character
di: Ma, Siyuan, et al.
Pubblicazione: (2024)