LM-DP-SGD estimates layer-specific MIA risks from shadow models and reweights gradients to give stronger protection to vulnerable layers, improving the privacy-utility trade-off over uniform DP-SGD.
Function- consistent feature distillation.arXiv preprint arXiv:2304.11832,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Mitigating Membership Inference in Intermediate Representations with Differentially Private Training
LM-DP-SGD estimates layer-specific MIA risks from shadow models and reweights gradients to give stronger protection to vulnerable layers, improving the privacy-utility trade-off over uniform DP-SGD.