Defending Against Saddle Point Attack in Byzantine-Robust Distributed Learning

Open in new window