no code implementations • 14 May 2024 • Jamie Heredge, Niraj Kumar, Dylan Herman, Shouvanik Chakrabarti, Romina Yalovetzky, Shree Hari Sureshbabu, Changhao Li, Marco Pistoia
We establish conditions on the encoding map such as classical simulatability, overlap with DLA basis, and its Fourier frequency characteristics that enable such a privacy breach of VQC models.
no code implementations • 22 Sep 2023 • Niraj Kumar, Jamie Heredge, Changhao Li, Shaltiel Eloul, Shree Hari Sureshbabu, Marco Pistoia
However, standard neural network-based federated learning models have been shown to be susceptible to data leakage from the gradients shared with the server.