# gradient-leakage

Federated Learning 攻擊s

攻擊ing federated learning through model update poisoning, gradient leakage, free-rider attacks, and Byzantine fault exploitation.