Approximate machine unlearning can raise the privacy risk of retained samples in differentially private models, according to a new augmentation-based membership inference audit.
Evaluations of machine learning privacy defenses are misleading
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.LG 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Auditing Approximate Machine Unlearning for Differentially Private Models
Approximate machine unlearning can raise the privacy risk of retained samples in differentially private models, according to a new augmentation-based membership inference audit.