mlsanityData to 14 Oct 2026

Improving User Privacy in Personalized Generation: Client-Side Retrieval-Augmented Modification of Server-Side Generated Speculations

arXiv 2601.17569 ↗First posted 24 Jan 2026cs.CL cs.AI cs.CRcited by 1

On the map: Retrieval-augmented generation → Privacy and security of RAG

Abstract (arXiv)

Personalization is crucial for aligning Large Language Model (LLM) outputs with individual user preferences and background knowledge. State-of-the-art solutions are based on retrieval augmentation, where relevant context from a user profile is retrieved for LLM consumption. These methods deal with a trade-off between exposing retrieved private data to cloud providers and relying on less capable local models. We introduce P3, an interactive framework for high-quality personalization without revealing private profiles to server-side LLMs. In P3, a large server-side model generates a sequence of k draft tokens based solely on the user query, while a small client-side model, with retrieval access to the user's private profile, evaluates and modifies these drafts to better reflect user preferences. This process repeats until an end token is generated. Experiments on LaMP-QA, a recent benchmark consisting of three personalized question answering datasets, show that P3 consistently outperforms both non-personalized server-side and personalized client-side baselines, achieving statistically significant improvements of 7.4% to 9% on average. Importantly, P3 recovers 90.3% to 95.7% of the utility of a ``leaky'' upper-bound scenario in which the full profile is exposed to the large server-side model. Privacy analyses, including linkability and attribute inference attacks, indicate that P3 preserves the privacy of a non-personalized server-side model, introducing only marginal additional leakage (1.5%--3.5%) compared to submitting a query without any personal context. Additionally, the framework is efficient for edge deployment, with the client-side model generating only 9.2% of the total tokens. These results demonstrate that P3 provides a practical, effective solution for personalized generation with improved privacy.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in Privacy and security of RAG, one of 15 directions in Retrieval-augmented generation. Builds on 35 of its 72 references that are in the map. First posted on arXiv 24 Jan 2026.

Builds on

35 of 72 references.

Closest ancestors · 1

Same direction, newest first

  1. 2024Mitigating the Privacy Issues in Retrieval-Augmented Generation (RAG) via Pure Synthetic Data20 Jun

Other references · 34

  1. 2023Efficient Memory Management for Large Language Model Serving with PagedAttentioncited by 5,944
  2. 2022Fast Inference from Transformers via Speculative Decodingcited by 1,653
  3. 2021Unsupervised Dense Information Retrieval with Contrastive Learningcited by 1,216
  4. 2023LaMP: When Large Language Models Meet Personalizationcited by 437
  5. 2023Personalized Soups: Personalized Large Language Model Alignment via Post-hoc Parameter Mergingcited by 280
  6. 2023Beyond Memorization: Violating Privacy Via Inference with Large Language Modelscited by 175
  7. 2024Personalized Language Modeling from Personalized Human Feedbackcited by 132
  8. 2024On-Device Language Models: A Comprehensive Reviewcited by 104
  9. 2024A Review on Edge Large Language Models: Design, Execution, and Applicationscited by 100
  10. 2024Optimization Methods for Personalizing Large Language Models through Retrieval Augmentationcited by 99
  11. 2024LongLaMP: A Benchmark for Personalized Long-form Text Generationcited by 97
  12. 2023Pearl: Personalizing Large Language Model Writing Assistants with Generation-Calibrated Retrieverscited by 92
22 more
  1. 2024Personalized Pieces: Efficient Personalized Large Language Models through Collaborative Effortscited by 74
  2. 2023LLM-based Medical Assistant Personalization with Short- and Long-Term Memory Coordinationcited by 46
  3. 2025Personalized Generation In Large Model Era: A Surveycited by 45
  4. 2023Learning to Rewrite Prompts for Personalized Text Generationcited by 40
  5. 2025LaMP-QA: A Benchmark for Personalized Long-form Question Answeringcited by 31
  6. 2024User Modeling and User Profiling: A Comprehensive Surveycited by 29
  7. 2025Reasoning-Enhanced Self-Training for Long-Form Personalized Text Generationcited by 27
  8. 2025A Survey of Personalization: From RAG to Agentcited by 27
  9. 2025ExPerT: Effective and Explainable Evaluation of Personalized Long-Form Text Generationcited by 26
  10. 2024Comparing Retrieval-Augmentation and Parameter-Efficient Fine-Tuning for Privacy-Preserving Personalization of Large Language Modelscited by 21
  11. 2017Measuring Personalization of Web Searchcited by 20
  12. 2025SoK: The Privacy Paradox of Large Language Models: Advancements, Privacy Risks, and Mitigationcited by 19
  13. 2024Corporate Communication Companion (CCC): An LLM-empowered Writing Assistant for Workplace Social Mediacited by 14
  14. 2024Learning to Rank for Multiple Retrieval-Augmented Models through Iterative Utility Maximizationcited by 14
  15. 2024Retrieval-Enhanced Machine Learning: Synthesis and Opportunitiescited by 11
  16. 2025Personalized Language Models via Privacy-Preserving Evolutionary Model Mergingcited by 8
  17. 2025Never Start from Scratch: Expediting On-Device LLM Personalization via Explainable Model Selectioncited by 8
  18. 2025PrLM: Learning Explicit Reasoning for Personalized RAG via Contrastive Reward Optimizationcited by 8
  19. 2025Learning from Natural Language Feedback for Personalized Question Answeringcited by 5
  20. 2025Pathways of Thoughts: Multi-Directional Thinking for Long-form Personalized Question Answeringcited by 5
  21. 2019Obfuscation for Privacy-preserving Syntactic Parsingcited by 3
  22. 2024Words Blending Boxes. Obfuscating Queries in Information Retrieval using Differential Privacycited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026PixARG: Pixel-Asynchronous Retrieval-Augmented Diffusion for Fast and Realistic Image Generationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Optimizing User Profiles via Contextual Bandits for Retrieval-Augmented LLM Personalization
  2. 2025Assessing RAG and HyDE on 1B vs. 4B-Parameter Gemma LLMs for Personal Assistants Integretion
  3. 2024Synthetic Data Generation with Large Language Models for Personalized Community Question Answering
  4. 2026SoK: Privacy Risks and Mitigations in Retrieval-Augmented Generation Systems
  5. 2026Learning to Refer: Client-Resolved Generation for Privacy-Aware Language Models
  6. 2026Less Is Personal: Learning Minimal Sufficient User Profiles for Personalized Language Models
  7. 2025Controlling What You Share: Assessing Language Model Adherence to Privacy Preferences
  8. 2025Privacy-Aware Decoding: Mitigating Privacy Leakage of Large Language Models in Retrieval-Augmented Generation