Illustration of the federated learning architecture showcasing the interaction between clients and a global server. Each client client_i computes a gradient ∇ω_i with respect to its local data and sends it to the global server. The server then aggregates these gradients (ω_g) to update the global model. The diagram also highlights a potential data leakage scenario, where the aggregated gradients can lead to the exposure of features from the real data, as depicted with the example images of a “cat” and a “dog”