Abstract Document Visual Question Answering (DocVQA) has quickly grown into a central
task of document understanding. But despite the fact that documents contain sensitive or …

Differentially private stochastic gradient descent (DP-SGD) is the standard algorithm for
training machine learning models under differential privacy (DP). The most common DP …

Linear models are ubiquitous in data science, but are particularly prone to overfitting and
data memorization in high dimensions. To guarantee the privacy of training data, differential …

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 …

Many high-stakes applications require machine learning models that protect user privacy
and provide well-calibrated, accurate predictions. While Differential Privacy (DP) is the gold …

We apply a state-of-the-art membership inference attack (MIA) to systematically test the
practical privacy vulnerability of fine-tuning large image classification models. We focus on …

The vision transformer-based foundation models, such as ViT or Dino-V2, are aimed at
solving problems with little or no finetuning of features. Using a setting of prototypical …

Transfer learning is a powerful technique that enables model training with limited amounts of
data, making it crucial in many data-scarce real-world applications. Typically, transfer …

Abstract Machine learning (ML) models have been shown to leak private information from
their training datasets. Differential Privacy (DP), typically implemented through the …

Federated learning (FL) is an emerging machine learning (ML) training paradigm where
clients own their data and collaborate to train a global model, without revealing any data to …