Cover Image

Through the Stealth Lens: Attention-Aware Defenses Against Poisoning in RAG

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Choudhary, Sarthak, Palumbo, Nils, Hooda, Ashish, Dvijotham, Krishnamurthy Dj, Jha, Somesh
Format:	Preprint
Published:	2025
Subjects:	Cryptography and Security Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

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

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

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

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

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

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

Hidden in the Metadata: Stealth Poisoning Attacks on Multimodal Retrieval-Augmented Generation
by: Edemacu, Kennedy, et al.
Published: (2026)

Semantic Chameleon: Corpus-Dependent Poisoning Attacks and Defenses in RAG Systems
by: Thornton, Scott
Published: (2026)

VeriGuard: Enhancing LLM Agent Safety via Verified Code Generation
by: Miculicich, Lesly, et al.
Published: (2025)

SuperLocalMemory: Privacy-Preserving Multi-Agent Memory with Bayesian Trust Defense Against Memory Poisoning
by: Bhardwaj, Varun Pratap
Published: (2026)

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

Defense Against Model Stealing Based on Account-Aware Distribution Discrepancy
by: Mei, Jian-Ping, et al.
Published: (2025)

Adversarial Machine Learning: Attacks, Defenses, and Open Challenges
by: Jha, Pranav K
Published: (2025)

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

ReliabilityRAG: Effective and Provably Robust Defense for RAG-based Web-Search
by: Shen, Zeyu, et al.
Published: (2025)

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

Dual Defense: Enhancing Privacy and Mitigating Poisoning Attacks in Federated Learning
by: Xu, Runhua, et al.
Published: (2025)

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

Behavior-Aware and Generalizable Defense Against Black-Box Adversarial Attacks for ML-Based IDS
by: Ennaji, Sabrine, et al.
Published: (2025)

The Defense Trilemma: Why Prompt Injection Defense Wrappers Fail?
by: Bhatt, Manish, et al.
Published: (2026)

Defending Against Beta Poisoning Attacks in Machine Learning Models
by: Gulciftci, Nilufer, et al.
Published: (2025)

Attacking Byzantine Robust Aggregation in High Dimensions
by: Choudhary, Sarthak, et al.
Published: (2023)

Malice in Agentland: Down the Rabbit Hole of Backdoors in the AI Supply Chain
by: Boisvert, Léo, et al.
Published: (2025)

From Poisoned to Aware: Fostering Backdoor Self-Awareness in LLMs
by: Shen, Guangyu, et al.
Published: (2025)

Cordon-MAS: Defending RAG against Knowledge Poisoning via Information-Flow Control
by: Yu, Zhe, et al.
Published: (2026)

FIDELIS: Blockchain-Enabled Protection Against Poisoning Attacks in Federated Learning
by: Carney, Jane, et al.
Published: (2025)

Phantom Transfer: Data-level Defences are Insufficient Against Data Poisoning
by: Draganov, Andrew, et al.
Published: (2026)

Class-Conditional Neural Polarizer: A Lightweight and Effective Backdoor Defense by Purifying Poisoned Features
by: Zhu, Mingli, et al.
Published: (2025)

P2P: A Poison-to-Poison Remedy for Reliable Backdoor Defense in LLMs
by: Zhao, Shuai, et al.
Published: (2025)

Defenses Against Prompt Attacks Learn Surface Heuristics
by: Li, Shawn, et al.
Published: (2026)

Stealthy Poisoning Attacks Bypass Defenses in Regression Settings
by: Carnerero-Cano, Javier, et al.
Published: (2026)

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

Enhancing Model Defense Against Jailbreaks with Proactive Safety Reasoning
by: Yang, Xianglin, et al.
Published: (2025)

Unraveling the Connections between Privacy and Certified Robustness in Federated Learning Against Poisoning Attacks
by: Xie, Chulin, et al.
Published: (2022)

SoK: Benchmarking Poisoning Attacks and Defenses in Federated Learning
by: Zhang, Heyi, et al.
Published: (2025)

Data to Defense: The Role of Curation in Customizing LLMs Against Jailbreaking Attacks
by: Liu, Xiaoqun, et al.
Published: (2024)

WARD: Adversarially Robust Defense of Web Agents Against Prompt Injections
by: Cao, Tri, et al.
Published: (2026)

WeiDetect: Weibull Distribution-Based Defense against Poisoning Attacks in Federated Learning for Network Intrusion Detection Systems
by: M., Sameera K., et al.
Published: (2025)

MELON: Provable Defense Against Indirect Prompt Injection Attacks in AI Agents
by: Zhu, Kaijie, et al.
Published: (2025)

DELMAN: Dynamic Defense Against Large Language Model Jailbreaking with Model Editing
by: Wang, Yi, et al.
Published: (2025)