Privacy Backdoors: Stealing Data with Corrupted Pretrained Models (Paper Explained)
Yannic Kilcher · 63:55 · Watch on YouTube
This video walks through the ETH Zurich paper "Privacy Backdoors: Stealing Data with Corrupted Pre-trained Models" by Shanglun Feng and Florian Tramèr. Its claim is that an attacker who publishes tampered pre-trained...