A malicious federated learning server can poison model weights so that all samples in a client batch become recoverable from gradients while evading the D-SNR detection metric.
R-gap: Recursive gradient attack on privacy,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CR 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
The Gradient Puppeteer: Adversarial Domination in Gradient Leakage Attacks through Model Poisoning
A malicious federated learning server can poison model weights so that all samples in a client batch become recoverable from gradients while evading the D-SNR detection metric.