142 citations · 842 across the 49 of their papers we have counts for
15 papers · 1 filter
Private Distribution Learning with Public Data: The View from Sample Compression
Shai Ben-David, Alex Bie, Clément L. Canonne +2
We study the problem of private distribution learning with access to public data. In this setup, which we refer to as public-private learning, the learner is given public and priva…
Exploring the Limits of Model-Targeted Indiscriminate Data Poisoning Attacks
Yiwei Lu, Gautam Kamath, Yaoliang Yu
Indiscriminate data poisoning attacks aim to decrease a model's test accuracy by injecting a small amount of corrupted training data. Despite significant interest, existing attacks…
Private GANs, Revisited
Alex Bie, Gautam Kamath, Guojun Zhang
We show that the canonical approach for training differentially private GANs -- updating the discriminator with differentially private stochastic gradient descent (DPSGD) -- can yi…
Hidden Poison: Machine Unlearning Enables Camouflaged Poisoning Attacks
Jimmy Z. Di, Jack Douglas, Jayadev Acharya +2
We introduce camouflaged data poisoning attacks, a new attack vector that arises in the context of machine unlearning and other settings when model retraining may be induced. An ad…
Position: Considerations for Differentially Private Learning with Large-Scale Public Pretraining
Florian Tramèr, Gautam Kamath, Nicholas Carlini
The performance of differentially private machine learning can be boosted significantly by leveraging the transfer learning capabilities of non-private models pretrained on large p…
Private Estimation with Public Data
Alex Bie, Gautam Kamath, Vikrant Singhal
We initiate the study of differentially private (DP) estimation with access to a small amount of public data. For private estimation of d-dimensional Gaussians, we assume that the…
Individual Privacy Accounting for Differentially Private Stochastic Gradient Descent
Da Yu, Gautam Kamath, Janardhan Kulkarni +3
Differentially private stochastic gradient descent (DP-SGD) is the workhorse algorithm for recent advances in private deep learning. It provides a single privacy guarantee to all d…
Indiscriminate Data Poisoning Attacks on Neural Networks
Yiwei Lu, Gautam Kamath, Yaoliang Yu
Data poisoning attacks, in which a malicious adversary aims to influence a model by injecting "poisoned" data into the training process, have attracted significant recent attention…
Differentially Private Fine-tuning of Language Models
Da Yu, Saurabh Naik, Arturs Backurs +9
We give simpler, sparser, and faster algorithms for differentially private fine-tuning of large-scale pre-trained language models, which achieve the state-of-the-art privacy versus…
Improved Rates for Differentially Private Stochastic Convex Optimization with Heavy-Tailed Data
Gautam Kamath, Xingtu Liu, Huanyu Zhang
We study stochastic convex optimization with heavy-tailed data under the constraint of differential privacy (DP). Most prior work on this problem is restricted to the case where th…
Remember What You Want to Forget: Algorithms for Machine Unlearning
Ayush Sekhari, Jayadev Acharya, Gautam Kamath +1
We study the problem of unlearning datapoints from a learnt model. The learner first receives a dataset drawn i.i.d. from an unknown distribution, and outputs a model $\widehat…
Enabling Fast Differentially Private SGD via Just-in-Time Compilation and Vectorization
Pranav Subramani, Nicholas Vadivelu, Gautam Kamath
A common pain point in differentially private machine learning is the significant runtime overhead incurred when executing Differentially Private Stochastic Gradient Descent (DPSGD…
Sever: A Robust Meta-Algorithm for Stochastic Optimization
Ilias Diakonikolas, Gautam Kamath, Daniel M. Kane +3
In high dimensions, most machine learning methods are brittle to even a small fraction of structured outliers. To address this, we introduce a new meta-algorithm that can take in a…
Actively Avoiding Nonsense in Generative Models
Steve Hanneke, Adam Kalai, Gautam Kamath +1
A generative model may generate utter nonsense when it is fit to maximize the likelihood of observed data. This happens due to "model error," i.e., when the true data generating di…
Being Robust (in High Dimensions) Can Be Practical
Ilias Diakonikolas, Gautam Kamath, Daniel M. Kane +3
Robust estimation is much more challenging in high dimensions than it is in one dimension: Most techniques either lead to intractable optimization problems or estimators that can t…