An important question in deploying large language models (LLMs) is how to augment LLMs with private data. We propose Differentially Private In-context Learning (DP-ICL) to enable LLMs to adapt to new tasks while maintaining privacy guarantees. DP-ICL performs private inference by establishing noisy consensus over an ensemble of exemplars using the Report-Noisy-Max mechanism. We evaluate DP-ICL on four benchmarks and find that it achieves comparable performance (<2\% degradation) with non-private ICL.

翻译：在部署大型语言模型（LLMs）时，一个重要问题是如何用私有数据增强LLMs。我们提出差分私有上下文学习（DP-ICL），使LLMs在保持隐私保证的同时适应新任务。DP-ICL通过使用汇报噪声最大值（Report-Noisy-Max）机制，在样本集合上建立噪声共识来实现私有推理。我们在四个基准上评估了DP-ICL，发现其性能与非私有ICL相当（性能下降<2%）。