Privacy-Preserving Retrieval Augmented Generation with Differential Privacy
- URL: http://arxiv.org/abs/2412.04697v1
- Date: Fri, 06 Dec 2024 01:20:16 GMT
- Title: Privacy-Preserving Retrieval Augmented Generation with Differential Privacy
- Authors: Tatsuki Koga, Ruihan Wu, Kamalika Chaudhuri,
- Abstract summary: Retrieval augmented generation (RAG) assists large language models (LLMs) by directly providing relevant information from external knowledge sources.
RAG outputs risk leaking sensitive information from the external data source.
In this work, we explore RAG under differential privacy (DP), a formal guarantee of data privacy.
- Score: 25.896416088293908
- License:
- Abstract: With the recent remarkable advancement of large language models (LLMs), there has been a growing interest in utilizing them in the domains with highly sensitive data that lies outside their training data. For this purpose, retrieval augmented generation (RAG) is particularly effective -- it assists LLMs by directly providing relevant information from the external knowledge sources. However, without extra privacy safeguards, RAG outputs risk leaking sensitive information from the external data source. In this work, we explore RAG under differential privacy (DP), a formal guarantee of data privacy. The main challenge with differentially private RAG is how to generate long accurate answers within a moderate privacy budget. We address this by proposing an algorithm that smartly spends privacy budget only for the tokens that require the sensitive information and uses the non-private LLM for other tokens. Our extensive empirical evaluations reveal that our algorithm outperforms the non-RAG baseline under a reasonable privacy budget of $\epsilon\approx 10$ across different models and datasets.
Related papers
- Calibrating Practical Privacy Risks for Differentially Private Machine Learning [5.363664265121231]
We study the approaches that can lower the attacking success rate to allow for more flexible privacy budget settings in model training.
We find that by selectively suppressing privacy-sensitive features, we can achieve lower ASR values without compromising application-specific data utility.
arXiv Detail & Related papers (2024-10-30T03:52:01Z) - The Good and The Bad: Exploring Privacy Issues in Retrieval-Augmented
Generation (RAG) [56.67603627046346]
Retrieval-augmented generation (RAG) is a powerful technique to facilitate language model with proprietary and private data.
In this work, we conduct empirical studies with novel attack methods, which demonstrate the vulnerability of RAG systems on leaking the private retrieval database.
arXiv Detail & Related papers (2024-02-23T18:35:15Z) - Making Differential Privacy Easier to Use for Data Controllers using a Privacy Risk Indicator [5.288762073608111]
Differential privacy (DP) enables private data analysis but is difficult to use in practice.
In a typical DP deployment, data controllers manage individuals' sensitive data and are responsible for answering data analysts' queries.
They do so by choosing $epsilon$, the privacy loss budget, which controls how much noise to add to the query output.
arXiv Detail & Related papers (2023-10-19T19:01:27Z) - Privacy Preserving Large Language Models: ChatGPT Case Study Based Vision and Framework [6.828884629694705]
This article proposes the conceptual model called PrivChatGPT, a privacy-generative model for LLMs.
PrivChatGPT consists of two main components i.e., preserving user privacy during the data curation/pre-processing together with preserving private context and the private training process for large-scale data.
arXiv Detail & Related papers (2023-10-19T06:55:13Z) - PrivacyMind: Large Language Models Can Be Contextual Privacy Protection Learners [81.571305826793]
We introduce Contextual Privacy Protection Language Models (PrivacyMind)
Our work offers a theoretical analysis for model design and benchmarks various techniques.
In particular, instruction tuning with both positive and negative examples stands out as a promising method.
arXiv Detail & Related papers (2023-10-03T22:37:01Z) - A Unified View of Differentially Private Deep Generative Modeling [60.72161965018005]
Data with privacy concerns comes with stringent regulations that frequently prohibited data access and data sharing.
Overcoming these obstacles is key for technological progress in many real-world application scenarios that involve privacy sensitive data.
Differentially private (DP) data publishing provides a compelling solution, where only a sanitized form of the data is publicly released.
arXiv Detail & Related papers (2023-09-27T14:38:16Z) - Privacy Implications of Retrieval-Based Language Models [26.87950501433784]
We present the first study of privacy risks in retrieval-based LMs, particularly $k$NN-LMs.
We find that $k$NN-LMs are more susceptible to leaking private information from their private datastore than parametric models.
arXiv Detail & Related papers (2023-05-24T08:37:27Z) - Algorithms with More Granular Differential Privacy Guarantees [65.3684804101664]
We consider partial differential privacy (DP), which allows quantifying the privacy guarantee on a per-attribute basis.
In this work, we study several basic data analysis and learning tasks, and design algorithms whose per-attribute privacy parameter is smaller that the best possible privacy parameter for the entire record of a person.
arXiv Detail & Related papers (2022-09-08T22:43:50Z) - Individual Privacy Accounting for Differentially Private Stochastic Gradient Descent [69.14164921515949]
We characterize privacy guarantees for individual examples when releasing models trained by DP-SGD.
We find that most examples enjoy stronger privacy guarantees than the worst-case bound.
This implies groups that are underserved in terms of model utility simultaneously experience weaker privacy guarantees.
arXiv Detail & Related papers (2022-06-06T13:49:37Z) - Private Reinforcement Learning with PAC and Regret Guarantees [69.4202374491817]
We design privacy preserving exploration policies for episodic reinforcement learning (RL)
We first provide a meaningful privacy formulation using the notion of joint differential privacy (JDP)
We then develop a private optimism-based learning algorithm that simultaneously achieves strong PAC and regret bounds, and enjoys a JDP guarantee.
arXiv Detail & Related papers (2020-09-18T20:18:35Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.