Mimic Function
Details about the Mimic attack method.
def mimic(learning_phase_mimic, update_mimic_heuristic, z_mimic, honest_vectors, current_step):
...
return honest_vectors[best_worker_to_mimic]
Inputs
learning_phase_mimic: bool - If True, the Mimic attack is applied.update_mimic_heuristic: Function - Function to update mu_mimic and z_mimic only if still in learning phasez_mimic: float - The z-score threshold for the Mimic heuristic.honest_vectors: list - List of honest vectors for processing.current_step: int - Current iteration or step number.
Outputs
torch.Tensor - The processed Tensor after applying the Mimic attack.