Privacy auditing with one (1) training run
We propose a scheme for auditing differentially private machine learning systems with a
single training run. This exploits the parallelism of being able to add or remove multiple …
single training run. This exploits the parallelism of being able to add or remove multiple …
Dp-forward: Fine-tuning and inference on language models with differential privacy in forward pass
Differentially private stochastic gradient descent (DP-SGD) adds noise to gradients in back-
propagation, safeguarding training data from privacy leakage, particularly membership …
propagation, safeguarding training data from privacy leakage, particularly membership …
One-shot empirical privacy estimation for federated learning
Privacy estimation techniques for differentially private (DP) algorithms are useful for
comparing against analytical bounds, or to empirically measure privacy loss in settings …
comparing against analytical bounds, or to empirically measure privacy loss in settings …
On differentially private federated linear contextual bandits
X Zhou, SR Chowdhury - arXiv preprint arXiv:2302.13945, 2023 - arxiv.org
We consider cross-silo federated linear contextual bandit (LCB) problem under differential
privacy, where multiple silos (agents) interact with the local users and communicate via a …
privacy, where multiple silos (agents) interact with the local users and communicate via a …
Practical differentially private hyperparameter tuning with subsampling
A Koskela, TD Kulkarni - Advances in Neural Information …, 2024 - proceedings.neurips.cc
Tuning the hyperparameters of differentially private (DP) machine learning (ML) algorithms
often requires use of sensitive data and this may leak private information via hyperparameter …
often requires use of sensitive data and this may leak private information via hyperparameter …
Efficient and near-optimal noise generation for streaming differential privacy
In the task of differentially private (DP) continual counting, we receive a stream of increments
and our goal is to output an approximate running total of these increments, without revealing …
and our goal is to output an approximate running total of these increments, without revealing …
Privacy amplification for matrix mechanisms
Privacy amplification exploits randomness in data selection to provide tighter differential
privacy (DP) guarantees. This analysis is key to DP-SGD's success in machine learning, but …
privacy (DP) guarantees. This analysis is key to DP-SGD's success in machine learning, but …
Subsampling suffices for adaptive data analysis
G Blanc - Proceedings of the 55th Annual ACM Symposium on …, 2023 - dl.acm.org
Ensuring that analyses performed on a dataset are representative of the entire population is
one of the central problems in statistics. Most classical techniques assume that the dataset is …
one of the central problems in statistics. Most classical techniques assume that the dataset is …
Privacy-Preserving Instructions for Aligning Large Language Models
Service providers of large language model (LLM) applications collect user instructions in the
wild and use them in further aligning LLMs with users' intentions. These instructions, which …
wild and use them in further aligning LLMs with users' intentions. These instructions, which …
Grafting Laplace and Gaussian distributions: A new noise mechanism for differential privacy
G Muthukrishnan, S Kalyani - IEEE Transactions on Information …, 2023 - ieeexplore.ieee.org
The framework of differential privacy protects an individual's privacy while publishing query
responses on congregated data. In this work, a new noise addition mechanism for …
responses on congregated data. In this work, a new noise addition mechanism for …