English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Recent advances in federated learning have demonstrated its promisingcapability to learn on decentralized datasets. However, a considerable amountof work has raised concerns due to the potential risks of adversariesparticipating in the framework to poison the global model for an adversarialpurpose. This paper investigates the feasibility of model poisoning forbackdoor attacks through rare word embeddings of NLP models. In textclassification, less than 1% of adversary clients suffices to manipulate themodel output without any drop in the performance on clean sentences. For a lesscomplex dataset, a mere 0.1% of adversary clients is enough to poison theglobal model effectively. We also propose a technique specialized in thefederated learning scheme called Gradient Ensemble, which enhances the backdoorperformance in all our experimental settings.

Backdoor Attacks in Federated Learning by Rare Embeddings and Gradient Ensembling