Differentially Private In-Context Learning with Nearest Neighbor Search
This addresses privacy concerns for users of large language models by providing a novel method to protect sensitive data during in-context learning, though it is incremental as it builds on existing DP-ICL approaches.
The paper tackles the privacy risks in in-context learning by introducing a differentially private framework that integrates nearest neighbor search, achieving more favorable privacy-utility trade-offs and outperforming existing baselines across benchmarks.
Differentially private in-context learning (DP-ICL) has recently become an active research topic due to the inherent privacy risks of in-context learning. However, existing approaches overlook a critical component of modern large language model (LLM) pipelines: the similarity search used to retrieve relevant context data. In this work, we introduce a DP framework for in-context learning that integrates nearest neighbor search of relevant examples in a privacy-aware manner. Our method outperforms existing baselines by a substantial margin across all evaluated benchmarks, achieving more favorable privacy-utility trade-offs. To achieve this, we employ nearest neighbor retrieval from a database of context data, combined with a privacy filter that tracks the cumulative privacy cost of selected samples to ensure adherence to a central differential privacy budget. Experimental results on text classification and document question answering show a clear advantage of the proposed method over existing baselines.