Abstract: Federated learning (FL), a distributed machine learning (ML) framework, is susceptible to Byzantine attacks since the attacker can manipulate clients’ local data or models to compromise the ...