all InfoSec news
SPEAR:Exact Gradient Inversion of Batches in Federated Learning
March 7, 2024, 5:11 a.m. | Dimitar I. Dimitrov, Maximilian Baader, Mark Niklas M\"uller, Martin Vechev
cs.CR updates on arXiv.org arxiv.org
Abstract: Federated learning is a popular framework for collaborative machine learning where multiple clients only share gradient updates on their local data with the server and not the actual data. Unfortunately, it was recently shown that gradient inversion attacks can reconstruct this data from these shared gradients. Existing attacks enable exact reconstruction only for a batch size of $b=1$ in the important honest-but-curious setting, with larger batches permitting only approximate reconstruction. In this work, we propose …
arxiv attacks can clients cs.cr cs.dc cs.lg data federated federated learning framework local machine machine learning popular server share updates
More from arxiv.org / cs.CR updates on arXiv.org
Jobs in InfoSec / Cybersecurity
Cyber Security Analyst
@ Dane Street | Palm Beach Gardens, Florida, United States
Program Information System Security Manager (ISSM) - onsite Tucson, AZ - TOP SECRET required
@ RTX | AZ855: RMS AP Bldg M05 1151 East Hermans Road Building M05, Tucson, AZ, 85756 USA
Lead - Business System Service (Workday HR Functional Consultant)
@ Freshworks | Bengaluru, India
Cloud Security Engineer
@ ButterflyMX | United States - Remote
Compliance Specialist
@ Airtable | Austin, Texas or San Francisco, California
Cyber SCRM Cloud Assessor Lead
@ Accenture Federal Services | Arlington, VA