DPNR preserves privacy of text representations using differential privacy.
problem Privacy leakage in deep learning text representations.
method DPNR uses Differential Privacy to provide formal privacy guarantees and dropout masking for enhanced privacy.
result DPNR reduces privacy leakage without significantly sacrificing main task performance.
Guaranteeing a certain level of user privacy in an arbitrary piece of text is a challenging issue. However, with this challenge comes the potential of unlocking access to vast data stores for training machine learning models and supporting data driven decisions. We address this problem through the lens of dx-privacy, a…
Generative text classifiers are most vulnerable to membership inference attacks.
problem Privacy threat from Membership Inference Attacks on generative text classifiers.
method Comprehensive empirical evaluation of generative, discriminative, and pseudo-generative classifiers across various datasets.
result Generative classifiers explicitly modeling P(X,Y) are most vulnerable to membership leakage. Proposes a text perturbation method using a Mahalanobis metric to balance privacy and utility.
problem Low utility of text analysis when using spherical noise for privacy-preserving text embedding.
method Regularized Mahalanobis metric to add elliptical noise, accounting for embedding space density.
result Improves privacy statistics while maintaining utility, outperforming Laplace mechanism.
Text-based analysis methods allow to reveal privacy relevant author attributes such as gender, age and identify of the text's author. Such methods can compromise the privacy of an anonymous author even when the author tries to remove privacy sensitive content. In this paper, we propose an automatic method, called Adver…
Develops a new deep learning framework for privacy-preserving text representations.
problem Privacy concerns in deep learning frameworks requiring data pooling to a trusted server.
method Three modules: embedding, randomization, and classifier. Novel LDP protocol reduces privacy impact on accuracy.
result Framework delivers comparable or better performance than non-private and existing LDP protocols.
Language models learn from training data and can leak private information.
problem Language models lack context understanding and can expose private data.
method Discussing the limitations of current privacy protection methods for language models.
result Existing privacy protection methods are insufficient for language models.
Framework extracts symptoms from EHRs for rapid disease outbreak detection.
problem Extracting relevant data from unstructured medical texts.
method Conformal active learning for efficient data mining.
result Framework achieves strong performance with minimal manual labeling.
New DP mechanism SWAG-PPM improves privacy in deep learning models.
problem Differential privacy struggles with real-world distributions, especially imbalanced data.
method SWAG-PPM uses a pseudo posterior distribution to downweight high-risk records.
result SWAG-PPM outperforms DP-SGD with similar privacy budget and modest utility degradation.
The Dirichlet mechanism protects privacy while minimizing KL divergence.
problem Minimizing KL divergence while protecting sensitive data privacy.
method Using the exponential mechanism with the KL divergence loss function, resulting in the Dirichlet mechanism.
result Proved a probability tail bound on KL divergence and derived a lower bound for sample complexity.
Accurately learning from user data while providing quantifiable privacy guarantees provides an opportunity to build better ML models while maintaining user trust. This paper presents a formal approach to carrying out privacy preserving text perturbation using the notion of dx-privacy designed to achieve geo-indistingui…
Latent Dirichlet Allocation (LDA) is a popular topic modeling technique for discovery of hidden semantic architecture of text datasets, and plays a fundamental role in many machine learning applications. However, like many other machine learning algorithms, the process of training a LDA model may leak the sensitive inf…
Paper improves deep learning privacy with new f-differential privacy framework.
problem Training neural networks on sensitive data while maintaining privacy.
method Introduced and analyzed f-differential privacy for neural networks training. result Improved privacy guarantees for neural networks training without sacrificing accuracy.
Dp-CLIP preserves privacy in multimodal AI training.
problem Privacy concerns in multimodal AI, especially in vision-language tasks.
method Differentially private adaptation of CLIP model.
result Dp-CLIP retains accuracy while ensuring privacy.
The Sampled Gaussian Mechanism's noise level decreases with larger subsampling rates, improving privacy-utility trade-offs.
problem Improving privacy-utility trade-offs in differentially private stochastic optimization.
method Proof of a conjecture about the Sampled Gaussian Mechanism's noise level and subsampling rate relationship.
result A rigorous proof of the conjecture, completing the proof of Theorem 6.2 in the original paper.
AdaDPS uses side information to improve private adaptive optimization.
problem Private adaptive optimization methods degrade when training with differential privacy.
method AdaDPS uses non-sensitive side information to precondition gradients.
result AdaDPS reduces the amount of noise needed for similar privacy guarantees, improving optimization performance.
The workshop focuses on AI principles for structured data.
problem Using AI on structured data for decision-making.
method Addressing principles of privacy, accountability, interpretability, robustness, and reasoning.
result Designing approaches to use structured data for reliable decisions.
We present a general method for privacy-preserving Bayesian inference in Poisson factorization, a broad class of models that includes some of the most widely used models in the social sciences. Our method satisfies limited precision local privacy, a generalization of local differential privacy, which we introduce to fo…
The goal of homomorphic encryption is to encrypt data such that another party can operate on it without being explicitly exposed to the content of the original data. We introduce an idea for a privacy-preserving transformation on natural language data, inspired by homomorphic encryption. Our primary tool is {\em obfusc…
New approach achieves optimal rates for differentially private stochastic convex optimization with heavy-tailed gradients.
problem Differentially private stochastic convex optimization with heavy-tailed gradients.
method Reduction-based approach to achieve optimal rates.
result Achieved optimal rates up to logarithmic factors, nearly matching a lower bound.
This paper presents a method to generate synthetic data with differential privacy to protect user privacy while maintaining data trends.
problem Protecting user privacy while using real data for AI applications.
method Develops a practical guide to generating synthetic data using differential privacy.
result Synthetic data can be generated to preserve trends of real data while ensuring strong privacy protections.
Generative AI tasks analyzed for text, images, audio, video, code, and molecules.
problem What is the core question when using generative AI?
method Survey of generative model families, probabilistic framework, game-theoretic setup, post-training modifications, socially responsible considerations.
result Generative AI is a distinct machine learning task with connections to prediction, compression, and decision-making.
Improved DMs with DP-SGD for generating private images.
problem Low privacy of DMs and lack of good privacy-utility tradeoff.
method Adopt LDMs with DP-SGD on attention modules of LDMs.
result Generates high-quality DP images with reduced parameters.
Federated Learning prioritizes client data contributions for better model quality.
problem Privacy concerns and reluctance to share private data in machine learning.
method Prioritizes client data contributions in Federated Learning by assigning scores based on defined criteria.
result The proposed approach yields a higher quality global model compared to standard Federated Learning.
New algorithm recovers matrices with unknown correspondences.
problem Recovering matrices from observations with unknown correspondences.
method Solves a nuclear norm minimization problem via proximal gradient with a Max-Oracle.
result Achieves state-of-the-art performance and high accuracy in recovering ground-truth correspondences.
Unified bounds for DP risks reduce noise and improve accuracy.
problem Difficult interpretation and calibration of DP mechanisms.
method Hypothesis-testing interpretation of DP (f-DP) and unified bounds. result Unified bounds are tighter and tunable for specific risks.
TextHide secures language understanding tasks by adding minimal encryption.
problem Mitigating privacy risks in natural language understanding tasks.
method TextHide adds a simple encryption step to prevent data recovery.
result TextHide reduces accuracy loss by only 1.9% on GLUE benchmark.
Differential privacy (DP) is a popular mechanism for training machine learning models with bounded leakage about the presence of specific points in the training data. The cost of differential privacy is a reduction in the model's accuracy. We demonstrate that in the neural networks trained using differentially private …
New tool for parallel and private stochastic convex optimization reduces query complexity.
problem Parallel and private stochastic convex optimization with reduced query complexity.
method Reweighted Stochastic Query (ReSQue) estimator combined with ball oracle acceleration.
result Achieves state-of-the-art complexities for SCO in parallel and private settings.
Improved privacy bounds for learning linear predictors with convex losses.
problem Differentially private learning of linear predictors with convex losses.
method Developed private model selection approach to achieve optimal rates.
result Improved upper and lower bounds for excess population risk.
Improved private AdaGrad achieves faster convergence rates for convex functions.
problem Private empirical risk minimization with differential privacy.
method Noisy AdaGrad with knowledge of gradient subspace geometry.
result Faster convergence rates for convex functions, bypassing traditional bounds.
Generative models help debug data issues in private, decentralized datasets.
problem Manual data inspection is impractical for private, decentralized datasets.
method Differentially private federated generative models.
result Generative models can effectively debug data issues in private datasets.
Study public-data assisted private stochastic optimization with labeled or unlabeled public data.
problem Limits and capability of public-data assisted differentially private (PA-DP) algorithms in stochastic convex optimization.
method Lower bounds for PA-DP mean estimation and novel methods for leveraging public data in private supervised learning.
result Achieved dimension independent rate for GLM with unlabeled public data, showing optimality.
Foundation models leak sensitive data in synthetic tabular data generation, especially LLaMA 3.3 70B.
problem Privacy leakage in synthetic tabular data generation using foundation models.
method Benchmarked three foundation models (GPT-4o-mini, LLaMA 3.3 70B, TabPFN v2) against four baselines on 35 real-world tables.
result Foundation models, especially LLaMA 3.3 70B, have the highest privacy risk in synthetic tabular data generation.
Paper accelerates and secures distributed NMF.
problem Efficiently processing large NMF matrices and maintaining data privacy.
method Proposes DSANLS framework with matrix sketching for acceleration and secure adaptation.
result DSANLS framework and secure adaptations for distributed NMF.
Privacy preserving networks can be modelled as decentralized networks (e.g., sensors, connected objects, smartphones), where communication between nodes of the network is not controlled by an all-knowing, central node. For this type of networks, the main issue is to gather/learn global information on the network (e.g.,…
Improves zeroth-order optimization for private machine learning with public data.
problem High computation and memory cost of first-order DP methods.
method PAZO (Public Data Assisted Zeroth-order Optimization) framework.
result Achieves superior privacy/utility tradeoffs across tasks.
Efficiently distills pretrained text-to-image models without real data, improving FID and CLIP scores.
problem Slow iterative refinement process of diffusion-based text-to-image models.
method Guided Score identity Distillation with Long and Short Classifier-Free Guidance.
result Achieves state-of-the-art FID performance with competitive CLIP score.
Statisticians contribute to LLMs for better trust and transparency.
problem Emerging statistical challenges in LLMs.
method Exploring statistical contributions to LLMs.
result Statisticians can enhance LLMs' trustworthiness and transparency.
Local adaptation improves federated learning models.
problem Improving accuracy of federated learning models on non-iid data.
method Local adaptation techniques (fine-tuning, multi-task learning, knowledge distillation).
result Participants benefit from local adaptation, improving federated model accuracy.
New algorithms for set union in differential privacy improve efficiency and accuracy.
problem Efficiently discovering items from private user data in natural language processing.
method Developed algorithms that allow users to contribute items in a dependent fashion, guided by a policy with contractive properties.
result New algorithms significantly outperform existing mechanisms in terms of efficiency and accuracy.
Study shows private learning of mixtures of Gaussians is possible with polynomial samples.
problem Estimating mixtures of Gaussians under differential privacy constraints.
method Developed a new framework for privately learning mixtures of Gaussians without structural assumptions.
result Polynomial number of samples (poly(k,d,1/α,1/ε,log(1/δ))) sufficient for estimation up to total variation distance α with (ε, δ)-DP.
Improves privacy amplification by shuffling for differential privacy.
problem Enhancing privacy guarantees in systems with anonymous data contributions.
method Theoretical and numerical analysis of Rényi differential privacy parameters and privacy amplification by shuffling.
result First asymptotically optimal analysis of Rényi differential privacy parameters for shuffled outputs.
Proposes a privacy-preserving recommendation system using matrix factorization and differential privacy.
problem Privacy leakage in recommendation systems when anonymizing user data is not sufficient.
method Uses matrix factorization and differential privacy via the Gaussian mechanism.
result Demonstrates excellent utility for privacy-preserving recommendation systems.
Federated learning with Bayesian differential privacy offers improved privacy and accuracy.
problem Privacy in federated learning with similar data distributions.
method Bayesian differential privacy for federated learning, with improved privacy budgeting.
result Significant advantage over state-of-the-art privacy bounds, with lower noise and improved accuracy.
This work optimizes mean estimation under varying privacy constraints.
problem Mean estimation with heterogeneous privacy constraints.
method Proposes an algorithm for mean estimation under different privacy levels for users.
result Shows a saturation phenomenon in performance as privacy levels are relaxed.
Improves privacy guarantees by analyzing randomness in privacy-preserving mechanisms.
problem Balancing user privacy and business constraints in privacy-preserving mechanisms.
method Analyzes explicit and implicit randomness in privacy mechanisms and proposes a probabilistic calibration method.
result Proposes privacy at risk, providing stronger privacy guarantees with quantifiable risks.
The paper analyzes and proposes methods for privately sharing individual privacy losses using per-instance differential privacy.
problem The standard differential privacy framework provides a worst-case bound that may not accurately reflect individual privacy losses.
method The paper analyzes per-instance differential privacy and proposes methods to privately and accurately publish per-instance privacy losses.
result The methods privately and accurately publish per-instance differential privacy losses with minimal additional privacy cost.