Fetching the paper…

Federated Learning with Extremely Noisy Clients via Negative Distillation · Around