Skip to main content
. 2023 May 22;14:2899. doi: 10.1038/s41467-023-38569-4

Fig. 3. Test performance with differential privacy (DP) training.

Fig. 3

Datasets are a MNIST, b Fashion-MNIST, and c CIFAR-10. Each figure reports mean and standard deviation over eight clients for each of five independent runs. Source data are provided as a Source Data file.