Imagen de Portada

Be Kind, Rewrite: Benign Projections via Rewriting Defend Against LLM Data Poisoning Attacks

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Halloran, John T., Bhatt, Noopur S.
Formato:	Preprint
Publicado:	2026
Materias:	Cryptography and Security Artificial Intelligence Machine Learning
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Defending Against Poisoning Attacks in Federated Learning with Blockchain
por: Dong, Nanqing, et al.
Publicado: (2023)

Have You Poisoned My Data? Defending Neural Networks against Data Poisoning
por: De Gaspari, Fabio, et al.
Publicado: (2024)

Defending Against Alignment-Breaking Attacks via Robustly Aligned LLM
por: Cao, Bochuan, et al.
Publicado: (2023)

Shadowcast: Stealthy Data Poisoning Attacks Against Vision-Language Models
por: Xu, Yuancheng, et al.
Publicado: (2024)

SPARD: Defending Harmful Fine-Tuning Attack via Safety Projection with Relevance-Diversity Data Selection
por: Chen, Shuhao, et al.
Publicado: (2026)

RL-Finetuned LLMs for Privacy-Preserving Synthetic Rewriting
por: Shi, Zhan, et al.
Publicado: (2025)

Revealing Weaknesses in Text Watermarking Through Self-Information Rewrite Attacks
por: Cheng, Yixin, et al.
Publicado: (2025)

Understanding the Effects of Safety Unalignment on Large Language Models
por: Halloran, John T.
Publicado: (2026)

Leveraging RAG for Training-Free Alignment of LLMs
por: Halloran, John T.
Publicado: (2026)

Efficient but Vulnerable: Benchmarking and Defending LLM Batch Prompting Attack
por: Yue, Murong, et al.
Publicado: (2025)

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

Attacks and Defenses Against LLM Fingerprinting
por: Kurian, Kevin, et al.
Publicado: (2025)

Data Poisoning Attacks on Off-Policy Policy Evaluation Methods
por: Lobo, Elita, et al.
Publicado: (2024)

MCP Safety Audit: LLMs with the Model Context Protocol Allow Major Security Exploits
por: Radosevich, Brandon, et al.
Publicado: (2025)

Defending Against Unforeseen Failure Modes with Latent Adversarial Training
por: Casper, Stephen, et al.
Publicado: (2024)

MCP Safety Training: Learning to Refuse Falsely Benign MCP Exploits using Improved Preference Alignment
por: Halloran, John
Publicado: (2025)

Defending Against Sophisticated Poisoning Attacks with RL-based Aggregation in Federated Learning
por: Wang, Yujing, et al.
Publicado: (2024)

Your Agent Can Defend Itself against Backdoor Attacks
por: Changjiang, Li, et al.
Publicado: (2025)

Rewriting the Budget: A General Framework for Black-Box Attacks Under Cost Asymmetry
por: Salmani, Mahdi, et al.
Publicado: (2025)

SAFELOC: Overcoming Data Poisoning Attacks in Heterogeneous Federated Machine Learning for Indoor Localization
por: Singampalli, Akhil, et al.
Publicado: (2024)

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

Learning to Defend by Attacking (and Vice-Versa): Transfer of Learning in Cybersecurity Games
por: Malloy, Tailia, et al.
Publicado: (2023)

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

Defending the Edge: Representative-Attention Defense against Backdoor Attacks in Federated Learning
por: Obioma, Chibueze Peace, et al.
Publicado: (2025)

MemoryGraft: Persistent Compromise of LLM Agents via Poisoned Experience Retrieval
por: Srivastava, Saksham Sahai, et al.
Publicado: (2025)

Machine Unlearning Fails to Remove Data Poisoning Attacks
por: Pawelczyk, Martin, et al.
Publicado: (2024)

GFCL: A GRU-based Federated Continual Learning Framework against Data Poisoning Attacks in IoV
por: Talpur, Anum, et al.
Publicado: (2022)

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

Scaling Trends for Data Poisoning in LLMs
por: Bowen, Dillon, et al.
Publicado: (2024)

Defending Against Weight-Poisoning Backdoor Attacks for Parameter-Efficient Fine-Tuning
por: Zhao, Shuai, et al.
Publicado: (2024)

XOXO: Stealthy Cross-Origin Context Poisoning Attacks against AI Coding Assistants
por: Štorek, Adam, et al.
Publicado: (2025)

Protecting Federated Learning from Extreme Model Poisoning Attacks via Multidimensional Time Series Anomaly Detection
por: Gabrielli, Edoardo, et al.
Publicado: (2023)

Best-of-Venom: Attacking RLHF by Injecting Poisoned Preference Data
por: Baumgärtner, Tim, et al.
Publicado: (2024)

Reasoning Introduces New Poisoning Attacks Yet Makes Them More Complicated
por: Foerster, Hanna, et al.
Publicado: (2025)

Universal Black-Box Reward Poisoning Attack against Offline Reinforcement Learning
por: Xu, Yinglun, et al.
Publicado: (2024)

ACE: A Model Poisoning Attack on Contribution Evaluation Methods in Federated Learning
por: Xu, Zhangchen, et al.
Publicado: (2024)

Hidden Poison: Machine Unlearning Enables Camouflaged Poisoning Attacks
por: Di, Jimmy Z., et al.
Publicado: (2022)

The Task Shield: Enforcing Task Alignment to Defend Against Indirect Prompt Injection in LLM Agents
por: Jia, Feiran, et al.
Publicado: (2024)

OEP: Poisoning Self-Evolving LLM Agents via Locally Correct but Non-Transferable Experiences
por: Wang, Kaixiang, et al.
Publicado: (2026)

PubDef: Defending Against Transfer Attacks From Public Models
por: Sitawarin, Chawin, et al.
Publicado: (2023)