Copertina

PRP: Propagating Universal Perturbations to Attack Large Language Model Guard-Rails

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Mangaokar, Neal, Hooda, Ashish, Choi, Jihye, Chandrashekaran, Shreyas, Fawaz, Kassem, Jha, Somesh, Prakash, Atul
Natura:	Preprint
Pubblicazione:	2024
Soggetti:	Cryptography and Security Computation and Language
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Defending Object Detectors against Patch Attacks with Out-of-Distribution Smoothing
di: Feng, Ryan, et al.
Pubblicazione: (2022)

What Really is a Member? Discrediting Membership Inference via Poisoning
di: Mangaokar, Neal, et al.
Pubblicazione: (2025)

PolicyLR: A Logic Representation For Privacy Policies
di: Hooda, Ashish, et al.
Pubblicazione: (2024)

Undetectable Backdoors in Model Parameters: Hiding Sparse Secrets in High Dimensions
di: Choudhary, Sarthak, et al.
Pubblicazione: (2026)

Functional Homotopy: Smoothing Discrete Optimization via Continuous Parameters for LLM Jailbreak Attacks
di: Wang, Zi, et al.
Pubblicazione: (2024)

Through the Stealth Lens: Attention-Aware Defenses Against Poisoning in RAG
di: Choudhary, Sarthak, et al.
Pubblicazione: (2025)

SEA: Shareable and Explainable Attribution for Query-based Black-box Attacks
di: Gao, Yue, et al.
Pubblicazione: (2023)

SLVR: Securely Leveraging Client Validation for Robust Federated Learning
di: Choi, Jihye, et al.
Pubblicazione: (2025)

Systems Security Foundations for Agentic Computing
di: Christodorescu, Mihai, et al.
Pubblicazione: (2025)

Taming Data Challenges in ML-based Security Tasks Using Generative AI
di: Kanchi, Shravya, et al.
Pubblicazione: (2025)

Harmful Terms and Where to Find Them: Measuring and Modeling Unfavorable Financial Terms and Conditions in Shopping Websites at Scale
di: Tsai, Elisa, et al.
Pubblicazione: (2025)

Agent Security is a Systems Problem
di: Christodorescu, Mihai, et al.
Pubblicazione: (2026)

Formal Policy Enforcement for Real-World Agentic Systems
di: Palumbo, Nils, et al.
Pubblicazione: (2026)

Prediction with Expert Advice under Local Differential Privacy
di: Jacobsen, Ben, et al.
Pubblicazione: (2025)

Private Continual Counting of Unbounded Streams
di: Jacobsen, Ben, et al.
Pubblicazione: (2025)

Confused ChatGPT: Cross-App Context Poisoning via First-Party APIs
di: Wang, Chao, et al.
Pubblicazione: (2026)

ARMOR: Aligning Secure and Safe Large Language Models via Meticulous Reasoning
di: Zhao, Zhengyue, et al.
Pubblicazione: (2025)

CIMemories: A Compositional Benchmark for Contextual Integrity of Persistent Memory in LLMs
di: Mireshghallah, Niloofar, et al.
Pubblicazione: (2025)

Fun-tuning: Characterizing the Vulnerability of Proprietary LLMs to Optimization-based Prompt Injection Attacks via the Fine-Tuning Interface
di: Labunets, Andrey, et al.
Pubblicazione: (2025)

Dependency-Aware Privacy for Multi-turn Agents
di: Anshumaan, Divyam, et al.
Pubblicazione: (2026)

JailGuard: A Universal Detection Framework for LLM Prompt-based Attacks
di: Zhang, Xiaoyu, et al.
Pubblicazione: (2023)

Personalizing Agent Privacy Decisions via Logical Entailment
di: Flemings, James, et al.
Pubblicazione: (2025)

SmoothGuard: Defending Multimodal Large Language Models with Noise Perturbation and Clustering Aggregation
di: Su, Guangzhi, et al.
Pubblicazione: (2025)

SmartGuard: Leveraging Large Language Models for Network Attack Detection through Audit Log Analysis and Summarization
di: Zhang, Hao, et al.
Pubblicazione: (2025)

Text-Based Personas for Simulating User Privacy Decisions
di: Fawaz, Kassem, et al.
Pubblicazione: (2026)

WebAgentGuard: A Reasoning-Driven Guard Model for Detecting Prompt Injection Attacks in Web Agents
di: Chen, Yulin, et al.
Pubblicazione: (2026)

Using Retriever Augmented Large Language Models for Attack Graph Generation
di: Prapty, Renascence Tarafder, et al.
Pubblicazione: (2024)

WebSP-Eval: Evaluating Web Agents on Website Security and Privacy Tasks
di: Ramesh, Guruprasad Viswanathan, et al.
Pubblicazione: (2026)

DualGuard: Dual-stream Large Language Model Watermarking Defense against Paraphrase and Spoofing Attack
di: Li, Hao, et al.
Pubblicazione: (2025)

Can One Safety Loop Guard Them All? Agentic Guard Rails for Federated Computing
di: Veeraragavan, Narasimha Raghavan, et al.
Pubblicazione: (2025)

TraceGuard: Process-Guided Firewall against Reasoning Backdoors in Large Language Models
di: Guo, Zhen, et al.
Pubblicazione: (2026)

Do Large Code Models Understand Programming Concepts? Counterfactual Analysis for Code Predicates
di: Hooda, Ashish, et al.
Pubblicazione: (2024)

SAGE: Sample-Aware Guarding Engine for Robust Intrusion Detection Against Adversarial Attacks
di: Chen, Jing, et al.
Pubblicazione: (2025)

KinGuard: Hierarchical Kinship-Aware Fingerprinting to Defend Against Large Language Model Stealing
di: Xu, Zhenhua, et al.
Pubblicazione: (2026)

ProP: Efficient Backdoor Detection via Propagation Perturbation for Overparametrized Models
di: Ren, Tao, et al.
Pubblicazione: (2024)

Publicly-Detectable Watermarking for Language Models
di: Fairoze, Jaiden, et al.
Pubblicazione: (2023)

How Not to Detect Prompt Injections with an LLM
di: Choudhary, Sarthak, et al.
Pubblicazione: (2025)

A New Era in LLM Security: Exploring Security Concerns in Real-World LLM-based Systems
di: Wu, Fangzhou, et al.
Pubblicazione: (2024)

On the Difficulty of Constructing a Robust and Publicly-Detectable Watermark
di: Fairoze, Jaiden, et al.
Pubblicazione: (2025)

Less Is More: Sparse and Cooperative Perturbation for Point Cloud Attacks
di: Tang, Keke, et al.
Pubblicazione: (2025)