Harnessing large-language models to generate private synthetic text Paper • 2306.01684 • Published Jun 2, 2023 • 3