322 citations · 945 across the 33 of their papers we have counts for
8 papers · 1 filter
Mixtral of Experts
Albert Q. Jiang, Alexandre Sablayrolles, Antoine Roux +23
We introduce Mixtral 8x7B, a Sparse Mixture of Experts (SMoE) language model. Mixtral has the same architecture as Mistral 7B, with the difference that each layer is composed of 8…
Privately generating tabular data using language models
Alexandre Sablayrolles, Yue Wang, Brian Karrer
Privately generating synthetic data from a table is an important brick of a privacy-first world. We propose and investigate a simple approach of treating each row in a table as a s…
Evaluating Privacy Leakage in Split Learning
Xinchi Qiu, Ilias Leontiadis, Luca Melis +2
Privacy-Preserving machine learning (PPML) can help us train and deploy models that utilize private information. In particular, on-device machine learning allows us to avoid sharin…
Analyzing Privacy Leakage in Machine Learning via Multiple Hypothesis Testing: A Lesson From Fano
Chuan Guo, Alexandre Sablayrolles, Maziar Sanjabi
Differential privacy (DP) is by far the most widely accepted framework for mitigating privacy risks in machine learning. However, exactly how small the privacy parameter needs…
TAN Without a Burn: Scaling Laws of DP-SGD
Tom Sander, Pierre Stock, Alexandre Sablayrolles
Differentially Private methods for training Deep Neural Networks (DNNs) have progressed recently, in particular with the use of massive batches and aggregated data augmentations fo…
CANIFE: Crafting Canaries for Empirical Privacy Measurement in Federated Learning
Samuel Maddock, Alexandre Sablayrolles, Pierre Stock
Federated Learning (FL) is a setting for training machine learning models in distributed environments where the clients do not share their raw data but instead send model updates t…
Defending against Reconstruction Attacks with Rényi Differential Privacy
Pierre Stock, Igor Shilov, Ilya Mironov +1
Reconstruction attacks allow an adversary to regenerate data samples of the training set using access to only a trained model. It has been recently shown that simple heuristics can…
Opacus: User-Friendly Differential Privacy Library in PyTorch
Ashkan Yousefpour, Igor Shilov, Alexandre Sablayrolles +9
We introduce Opacus, a free, open-source PyTorch library for training deep learning models with differential privacy (hosted at opacus.ai). Opacus is designed for simplicity, flexi…