More Like this

1. Private selection from private candidates

   https://doi.org/10.1145/3313276.3316377  

   Liu, Jingcheng; Talwar, Kunal (June 2019, Proceedings of the ACM Symposium on Theory of Computing (STOC))
2. Auditing Differentially Private Machine Learning: How Private is Private SGD?

   Jagielski, Matthew; Ullman, Jonathan; Oprea, Alina (January 2020, Advances in Neural and Information Processing Systems)

   null (Ed.)
3. When is Differentially Private Finetuning Private?

   Rinberg, Roy; Pawelczyk, Martin (October 2024, Statistical Foundations of LLMs and Foundation Models (NeurIPS 2024 Workshop))

   Differential Privacy (DP) is a mathematical definition that enshrines a formal guarantee that the output of a query does not depend greatly on any individual in the dataset. DP does not formalize a notion of "background information" and does not provide a guarantee about how much an output can be identifying to someone who has background information about an individual. In this paper, we argue that privately fine-tuning a pre-trained machine learning model on a private dataset using differential privacy does not always yield meaningful notions of privacy. Simply offering differential privacy guarantees in terms of (ε, δ) is insufficient to ensure human notions privacy, when the original training data is correlated with the fine-tuning dataset. We emphasize that, alongside differential privacy assurances, it is essential to report measures of dataset similarity and model attackability (for which model-size can be a proxy). This is a work in progress; this work is primarily a position piece, arguing for how DP should be used in practice, and what future research needs to be conducted in order to better answer those questions. 

   more » « less
4. Cheaper Private Set Intersection via Differentially Private Leakage

   https://doi.org/10.2478/popets-2019-0034  

   Groce, Adam; Rindal, Peter; Rosulek, Mike (July 2019, Proceedings on Privacy Enhancing Technologies)

   Abstract In this work we demonstrate that allowing differentially private leakage can significantly improve the concrete performance of secure 2-party computation (2PC) protocols. Specifically, we focus on the private set intersection (PSI) protocol of Rindal and Rosulek (CCS 2017), which is the fastest PSI protocol with security against malicious participants. We show that if differentially private leakage is allowed, the cost of the protocol can be reduced by up to 63%, depending on the desired level of differential privacy. On the technical side, we introduce a security model for differentially-private leakage in malicious-secure 2PC. We also introduce two new and improved mechanisms for “differentially private histogram overestimates,” the main technical challenge for differentially-private PSI. 

   more » « less
5. Private Information Retrieval With Private Noisy Side Information

   https://doi.org/10.1109/TIT.2024.3363133  

   ZivariFard, Hassan; Chou, Rémi A. (April 2024, IEEE Transactions on Information Theory)