5 papers
One Pic is All it Takes: Poisoning Visual Document Retrieval Augmented Generation with a Single Image
Ezzeldin Shereen, Dan Ristea, Shae McFadden +3
Retrieval-augmented generation (RAG) is instrumental for inhibiting hallucinations in large language models (LLMs) through the use of a factual knowledge base (KB). Although PDF do…
Poisoning Attacks on LLMs Require a Near-constant Number of Poison Samples
Alexandra Souly, Javier Rando, Ed Chapman +10
Poisoning attacks can compromise the safety of large language models (LLMs) by injecting malicious documents into their training data. Existing work has studied pretraining poisoni…
From Promise to Peril: Rethinking Cybersecurity Red and Blue Teaming in the Age of LLMs
Alsharif Abuadbba, Chris Hicks, Kristen Moore +4
Large Language Models (LLMs) are set to reshape cybersecurity by augmenting red and blue team operations. Red teams can exploit LLMs to plan attacks, craft phishing content, simula…
Private Collaborative Edge Inference via Over-the-Air Computation
Selim F. Yilmaz, Burak Hasircioglu, Li Qiao +1
We consider collaborative inference at the wireless edge, where each client's model is trained independently on its local dataset. Clients are queried in parallel to make an accura…
Generalized Multivariate Polynomial Codes for Distributed Matrix-Matrix Multiplication
Jesús Gómez-Vilardebó, Burak HasırcıoÄlu, Deniz Gündüz
Supporting multiple partial computations efficiently at each of the workers is a keystone in distributed coded computing in order to speed up computations and to fully exploit the…