Popular GitHub Repositories for Poison

The official implementation of the CCS'23 paper, Narcissus clean-label backdoor attack -- only takes THREE images to poison a face recognition dataset in a clean-label way and achieves a 99.89% attack success rate.

A Survey of Poisoning Attacks and Defenses in Recommender Systems

Poison as Cure: Visual Noise for Mitigating Object Hallucinations in LVMs

DIVA -- Poison is not Traceless: Black-Box Detection of Poisoning Attacks

History Poison Lab: Vulnerable LLM implementation demonstrating Chat History Poisoning attacks. Learn how attackers manipulate chat context and explore mitigation strategies for secure LLM applications.