mlsanityData to 7 Oct 2026

Differentially Private Datastore Generation for Retrieval-Augmented Inference

arXiv 2606.01413 ↗First posted 31 May 2026cs.CR cs.IR cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

It is crucial for modern on-device AI systems that rely on retrieval-augmented inference to release and share datastores without compromising individual privacy. This can be achieved using Differential Privacy (DP), which provides a formal guarantee that ensures individual contributions remain indistinguishable, even under adversarial analysis. In this paper, we introduce a hashing-based probability generation framework designed to enable the creation and release of differentially private datastores. Our approach employs locality-sensitive hashing (LSH) to efficiently partition high-dimensional data into buckets. We then add calibrated DP noise to the accumulated vote for each bucket, generating a probability distribution across classes. Our method is broadly applicable to any pipeline requiring secure key,value datastore creation and release. We conducted experiments on seven datasets with varying sample sizes and class counts, ranging from 2 to 14. At epsilon=5, our released DP datastore achieves strong privacy protection with only an average 2.6% drop in accuracy. Finally, we benchmark DP datastore resilience to membership inference attacks, reducing attack accuracy to 53.60%.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 17 of its 32 references that are in the map. First posted on arXiv 31 May 2026.

Builds on

17 of 32 references.

References · 17

  1. 2020Language Models are Few-Shot Learnerscited by 32,199
  2. 2016Deep Learning with Differential Privacycited by 2,682
  3. 2015Character-level Convolutional Networks for Text Classificationcited by 2,293
  4. 2016Membership Inference Attacks against Machine Learning Modelscited by 2,221
  5. 2019Generalization through Memorization: Nearest Neighbor Language Modelscited by 788
  6. 2025Phi-4-Mini Technical Report: Compact yet Powerful Multimodal Language Models via Mixture-of-LoRAscited by 491
  7. 2020Nearest Neighbor Machine Translationcited by 209
  8. 2023How to DP-fy ML: A Practical Guide to Machine Learning with Differential Privacycited by 150
  9. 2023Privacy-Preserving In-Context Learning with Differentially Private Few-Shot Generationcited by 94
  10. 2023Privacy-Preserving In-Context Learning for Large Language Modelscited by 50
  11. 2023kNN Prompting: Beyond-Context Learning with Calibration-Free Nearest Neighbor Inferencecited by 47
  12. 2025A Survey on Knowledge-Oriented Retrieval-Augmented Generationcited by 34
5 more
  1. 2022DP-Rewrite: Towards Reproducibility and Transparency in Differentially Private Text Rewritingcited by 28
  2. 2023DP-BART for Privatized Text Rewriting under Local Differential Privacycited by 27
  3. 2023Simple and Scalable Nearest Neighbor Machine Translationcited by 12
  4. 2023Locality-Sensitive Hashing Does Not Guarantee Privacy! Attacks on Google's FLoC and the MinHash Hierarchy Systemcited by 1
  5. 2024Differentially Private High-Dimensional Approximate Range Counting, Revisitedcited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2024Privacy-Preserving Retrieval-Augmented Generation with Differential Privacy
  2. 2026Efficient Privacy-Preserving Retrieval Augmented Generation with Distance-Preserving Encryption
  3. 2026Pointing the Way, Hiding the Destination: Practical Private Dense Retrieval at Scale
  4. 2024RemoteRAG: A Privacy-Preserving LLM Cloud RAG Service
  5. 2025RAPID: Retrieval Augmented Training of Differentially Private Diffusion Models
  6. 2025Towards Confidential and Efficient LLM Inference with Dual Privacy Protection
  7. 2026Differentially Private Retrieval-Augmented Generation
  8. 2026Quantization Enables Private Dense Retrieval against Malicious Service Providers