Copertina

BodhiPromptShield: Pre-Inference Prompt Mediation for Suppressing Privacy Propagation in LLM/VLM Agents

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Ma, Bo, Wu, Jinsong, Yan, Weiqi
Natura:	Preprint
Pubblicazione:	2026
Soggetti:	Cryptography and Security Computer Vision and Pattern Recognition
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Bodhi VLM: Privacy-Alignment Modeling for Hierarchical Visual Representations in Vision Backbones and VLM Encoders via Bottom-Up and Top-Down Feature Search
di: Ma, Bo, et al.
Pubblicazione: (2026)

REAEDP: Entropy-Calibrated Differentially Private Data Release with Formal Guarantees and Attack-Based Evaluation
di: Ma, Bo, et al.
Pubblicazione: (2026)

UnlearnShield: Shielding Forgotten Privacy against Unlearning Inversion
di: Xue, Lulu, et al.
Pubblicazione: (2026)

Are You Copying My Prompt? Protecting the Copyright of Vision Prompt for VPaaS via Watermark
di: Ren, Huali, et al.
Pubblicazione: (2024)

Who Can See Through You? Adversarial Shielding Against VLM-Based Attribute Inference Attacks
di: Fan, Yucheng, et al.
Pubblicazione: (2025)

PromptSmooth: Certifying Robustness of Medical Vision-Language Models via Prompt Learning
di: Hussein, Noor, et al.
Pubblicazione: (2024)

AutoMIA: Improved Baselines for Membership Inference Attack via Agentic Self-Exploration
di: Liu, Ruhao, et al.
Pubblicazione: (2026)

A Cross-Modal Prompt Injection Attack against Large Vision-Language Models with Image-Only Perturbation
di: Yang, Hao, et al.
Pubblicazione: (2026)

EO-VLM: VLM-Guided Energy Overload Attacks on Vision Models
di: Seo, Minjae, et al.
Pubblicazione: (2025)

SecureT2I: No More Unauthorized Manipulation on AI Generated Images from Prompts
di: Wu, Xiaodong, et al.
Pubblicazione: (2025)

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

Backdoor Attacks on Prompt-Driven Video Segmentation Foundation Models
di: Zhang, Zongmin, et al.
Pubblicazione: (2025)

Jailbreak Vision Language Models via Bi-Modal Adversarial Prompt
di: Ying, Zonghao, et al.
Pubblicazione: (2024)

BadBone: Backdoor Attacks Against Backbone Models in Visual Prompt Learning
di: Yang, Ziqing, et al.
Pubblicazione: (2026)

Beyond Text Prompts: Precise Concept Erasure through Text-Image Collaboration
di: Li, Jun, et al.
Pubblicazione: (2026)

SPARK: Jailbreaking T2V Models by Synergistically Prompting Auditory and Recontextualized Knowledge
di: Ying, Zonghao, et al.
Pubblicazione: (2025)

SurrogatePrompt: Bypassing the Safety Filter of Text-to-Image Models via Substitution
di: Ba, Zhongjie, et al.
Pubblicazione: (2023)

Not All Prompts Are Secure: A Switchable Backdoor Attack Against Pre-trained Vision Transformers
di: Yang, Sheng, et al.
Pubblicazione: (2024)

Exposing Functional Fusion: A New Class of Strategic Backdoor in Dynamic Prompt Architectures
di: Liu, Zeyao, et al.
Pubblicazione: (2026)

Disentangling Adversarial Prompts: A Semantic-Graph Defense for Robust LLM Security
di: Fang, Xiang, et al.
Pubblicazione: (2026)

Improving Adversarial Transferability on Vision Transformers via Forward Propagation Refinement
di: Ren, Yuchen, et al.
Pubblicazione: (2025)

ComPrivDet: Efficient Privacy Object Detection in Compressed Domains Through Inference Reuse
di: Yao, Yunhao, et al.
Pubblicazione: (2026)

Volley Revolver: A Novel Matrix-Encoding Method for Privacy-Preserving Neural Networks (Inference)
di: Chiang, John
Pubblicazione: (2022)

Mind the Third Eye! Benchmarking Privacy Awareness in MLLM-powered Smartphone Agents
di: Lin, Zhixin, et al.
Pubblicazione: (2025)

DAVSP: Safety Alignment for Large Vision-Language Models via Deep Aligned Visual Safety Prompt
di: Zhang, Yitong, et al.
Pubblicazione: (2025)

Revisiting the Privacy Risks of Split Inference: A GAN-Based Data Reconstruction Attack via Progressive Feature Optimization
di: Qiu, Yixiang, et al.
Pubblicazione: (2025)

Image-based Prompt Injection: Hijacking Multimodal LLMs through Visually Embedded Adversarial Instructions
di: Nagaraja, Neha, et al.
Pubblicazione: (2026)

One Prompt to Verify Your Models: Black-Box Text-to-Image Models Verification via Non-Transferable Adversarial Attacks
di: Guo, Ji, et al.
Pubblicazione: (2024)

SlowBA: An efficiency backdoor attack towards VLM-based GUI agents
di: Li, Junxian, et al.
Pubblicazione: (2026)

EditShield: Protecting Unauthorized Image Editing by Instruction-guided Diffusion Models
di: Chen, Ruoxi, et al.
Pubblicazione: (2023)

Iteratively Prompting Multimodal LLMs to Reproduce Natural and AI-Generated Images
di: Naseh, Ali, et al.
Pubblicazione: (2024)

Few-Shot Adversarial Prompt Learning on Vision-Language Models
di: Zhou, Yiwei, et al.
Pubblicazione: (2024)

FT-Shield: A Watermark Against Unauthorized Fine-tuning in Text-to-Image Diffusion Models
di: Cui, Yingqian, et al.
Pubblicazione: (2023)

BSPA: Exploring Black-box Stealthy Prompt Attacks against Image Generators
di: Tian, Yu, et al.
Pubblicazione: (2024)

Harnessing Hyperbolic Geometry for Harmful Prompt Detection and Sanitization
di: Maljkovic, Igor, et al.
Pubblicazione: (2026)

GEO-Detective: Unveiling Location Privacy Risks in Images with LLM Agents
di: Zhang, Xinyu, et al.
Pubblicazione: (2025)

Be Careful What You Smooth For: Label Smoothing Can Be a Privacy Shield but Also a Catalyst for Model Inversion Attacks
di: Struppek, Lukas, et al.
Pubblicazione: (2023)

VLATTACK: Multimodal Adversarial Attacks on Vision-Language Tasks via Pre-trained Models
di: Yin, Ziyi, et al.
Pubblicazione: (2023)

PLA: Prompt Learning Attack against Text-to-Image Generative Models
di: Lyu, Xinqi, et al.
Pubblicazione: (2025)

Fooling the Watchers: Breaking AIGC Detectors via Semantic Prompt Attacks
di: Hao, Run, et al.
Pubblicazione: (2025)