Marco Mondelli
@mmondelli.bsky.social
Professor at IST Austria | Machine learning, information theory | Prev: Stanford, EPFL | opinions my own
Reposted by Marco Mondelli
“Privacy for free”: On deep learning, model size, and performance cost
Can privacy be guaranteed when deep learning training models become larger? Does this come with higher performance costs?
Can privacy be guaranteed when deep learning training models become larger? Does this come with higher performance costs?
April 15, 2025 at 10:37 AM
“Privacy for free”: On deep learning, model size, and performance cost
Can privacy be guaranteed when deep learning training models become larger? Does this come with higher performance costs?
Can privacy be guaranteed when deep learning training models become larger? Does this come with higher performance costs?