PRoVeFL Framework Enables Verifiable and Byzantine-Robust Federated Learning

PRoVeFL utilizes multi-key fully homomorphic encryption across multiple servers to provide privacy-preserving, Byzantine-robust, and verifiable aggregation. This modular framework addresses the trade-off between privacy, integrity, and computational overhead in decentralized training.