315 citations · 790 across the 31 of their papers we have counts for
33 papers
AfterDeleteBench: Auditing Post-Deletion Information Persistence in LLM Agents
Ehsan Saleh, Roozbeh Nahavandi, Sama Salarian +2
Many LLM agents maintain external persistent memory and derive additional state from past interactions, such as summaries. Deleting an original memory record therefore need not rem…
Data Sharing with Endogenous Choices over Differential Privacy Levels
Raef Bassily, Kate Donahue, Diptangshu Sen +2
Motivated by the rapid push to decentralize sharing of data, we study whether large-scale data sharing coalitions can form in a decentralized manner under differential privacy when…
Private Model Personalization Revisited
Conor Snedeker, Xinyu Zhou, Raef Bassily
We study model personalization under user-level differential privacy (DP) in the shared representation framework. In this problem, there are users whose data is statistically h…
Private Algorithms for Stochastic Saddle Points and Variational Inequalities: Beyond Euclidean Geometry
Raef Bassily, Cristóbal Guzmán, Michael Menart
In this work, we conduct a systematic study of stochastic saddle point problems (SSP) and stochastic variational inequalities (SVI) under the constraint of -differential pri…
Public-data Assisted Private Stochastic Optimization: Power and Limitations
Enayat Ullah, Michael Menart, Raef Bassily +2
We study the limits and capability of public-data assisted differentially private (PA-DP) algorithms. Specifically, we focus on the problem of stochastic convex optimization (SCO)…
Differentially Private Worst-group Risk Minimization
Xinyu Zhou, Raef Bassily
We initiate a systematic study of worst-group risk minimization under -differential privacy (DP). The goal is to privately find a model that approximately minimizes the max…
Differentially Private Non-Convex Optimization under the KL Condition with Optimal Rates
Michael Menart, Enayat Ullah, Raman Arora +2
We study private empirical risk minimization (ERM) problem for losses satisfying the -Kurdyka-Łojasiewicz (KL) condition. The Polyak-Łojasiewicz (PL) condition is a special…
Differentially Private Domain Adaptation with Theoretical Guarantees
Raef Bassily, Corinna Cortes, Anqi Mao +1
In many applications, the labeled data at the learner's disposal is subject to privacy constraints and is relatively limited. To derive a more accurate predictor for the target dom…
Differentially Private Algorithms for the Stochastic Saddle Point Problem with Optimal Rates for the Strong Gap
Raef Bassily, Cristóbal Guzmán, Michael Menart
We show that convex-concave Lipschitz stochastic saddle point problems (also known as stochastic minimax optimization) can be solved under the constraint of -differential pr…
Private Domain Adaptation from a Public Source
Raef Bassily, Mehryar Mohri, Ananda Theertha Suresh
A key problem in a variety of applications is that of domain adaptation from a public source domain, for which a relatively large amount of labeled data with no privacy constraints…
Faster Rates of Convergence to Stationary Points in Differentially Private Optimization
Raman Arora, Raef Bassily, Tomás González +3
We study the problem of approximating stationary points of Lipschitz and smooth functions under -differential privacy (DP) in both the finite-sum and stochastic se…
Differentially Private Generalized Linear Models Revisited
Raman Arora, Raef Bassily, Cristóbal Guzmán +2
We study the problem of -differentially private learning of linear predictors with convex losses. We provide results for two subclasses of loss functions. The first case is…
Differentially Private Learning with Margin Guarantees
Raef Bassily, Mehryar Mohri, Ananda Theertha Suresh
We present a series of new differentially private (DP) algorithms with dimension-independent margin guarantees. For the family of linear hypotheses, we give a pure DP learning algo…
Differentially Private Stochastic Optimization: New Results in Convex and Non-Convex Settings
Raef Bassily, Cristóbal Guzmán, Michael Menart
We study differentially private stochastic optimization in convex and non-convex settings. For the convex case, we focus on the family of non-smooth generalized linear losses (GLLs…
Non-Euclidean Differentially Private Stochastic Convex Optimization: Optimal Rates in Linear Time
Raef Bassily, Cristóbal Guzmán, Anupama Nandi
Differentially private (DP) stochastic convex optimization (SCO) is a fundamental problem, where the goal is to approximately minimize the population risk with respect to a convex…
Learning from Mixtures of Private and Public Populations
Raef Bassily, Shay Moran, Anupama Nandi
We initiate the study of a new model of supervised learning under privacy constraints. Imagine a medical study where a dataset is sampled from a population of both healthy and unhe…
Stability of Stochastic Gradient Descent on Nonsmooth Convex Losses
Raef Bassily, Vitaly Feldman, Cristóbal Guzmán +1
Uniform stability is a notion of algorithmic stability that bounds the worst case change in the model output by the algorithm when a single data point in the dataset is replaced. A…
Private Query Release Assisted by Public Data
Raef Bassily, Albert Cheu, Shay Moran +3
We study the problem of differentially private query release assisted by access to public data. In this problem, the goal is to answer a large class of statistical qu…
Limits of Private Learning with Access to Public Data
Noga Alon, Raef Bassily, Shay Moran
We consider learning problems where the training set consists of two types of examples: private and public. The goal is to design a learning algorithm that satisfies differential p…
Private Stochastic Convex Optimization with Optimal Rates
Raef Bassily, Vitaly Feldman, Kunal Talwar +1
We study differentially private (DP) algorithms for stochastic convex optimization (SCO). In this problem the goal is to approximately minimize the population loss given i.i.d. sam…
Privately Answering Classification Queries in the Agnostic PAC Model
Anupama Nandi, Raef Bassily
We revisit the problem of differentially private release of classification queries. In this problem, the goal is to design an algorithm that can accurately answer a sequence of cla…
On exponential convergence of SGD in non-convex over-parametrized learning
Raef Bassily, Mikhail Belkin, Siyuan Ma
Large over-parametrized models learned via stochastic gradient descent (SGD) methods have become a key element in modern machine learning. Although SGD methods are very effective i…
Linear Queries Estimation with Local Differential Privacy
Raef Bassily
We study the problem of estimating a set of linear queries with respect to some unknown distribution over a domain based on a sensitive data set…
Model-Agnostic Private Learning via Stability
Raef Bassily, Om Thakkar, Abhradeep Thakurta
We design differentially private learning algorithms that are agnostic to the learning model. Our algorithms are interactive in nature, i.e., instead of outputting a model based on…