Cross-silo federated learning (FL) enables organizations (e.g., financial or medical) to collaboratively train a machine learning model by aggregating local gradient updates from each client without sharing privacy-sensitive data. To ensure no update …