could we support Gaussian/Laplace Differential Privacy on homoxgb histgram exchange? #4866
Unanswered
EasonFeng5870
asked this question in
Q&A
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
hi there.
homo-sbt is using dh for random number generator seed. but there is paper <> show that gradients reveal some properties of the training data, also show the Defense Strategies, like Noisy Gradients or Gradient Compression and Sparsification or cryptography.
Is there a way we could support them?
Beta Was this translation helpful? Give feedback.
All reactions