Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox

Picture for Thomas Steinke

PAC-Bayes, MAC-Bayes and Conditional Mutual Information: Fast rate bounds that handle general VC classes


Jun 17, 2021
Peter Grünwald, Thomas Steinke, Lydia Zakynthinou

* 24 pages, accepted for publication at COLT 2021 

  Access Paper or Ask Questions

Privately Learning Subspaces


May 28, 2021
Vikrant Singhal, Thomas Steinke


  Access Paper or Ask Questions

Leveraging Public Data for Practical Private Query Release


Feb 17, 2021
Terrance Liu, Giuseppe Vietri, Thomas Steinke, Jonathan Ullman, Zhiwei Steven Wu


  Access Paper or Ask Questions

The Distributed Discrete Gaussian Mechanism for Federated Learning with Secure Aggregation


Feb 12, 2021
Peter Kairouz, Ziyu Liu, Thomas Steinke


  Access Paper or Ask Questions

New Oracle-Efficient Algorithms for Private Synthetic Data Release


Jul 10, 2020
Giuseppe Vietri, Grace Tian, Mark Bun, Thomas Steinke, Zhiwei Steven Wu


  Access Paper or Ask Questions

The Discrete Gaussian for Differential Privacy


Apr 07, 2020
Clément Canonne, Gautam Kamath, Thomas Steinke

* Working paper, comments welcome 

  Access Paper or Ask Questions

Reasoning About Generalization via Conditional Mutual Information


Feb 14, 2020
Thomas Steinke, Lydia Zakynthinou

* 56 pages 

  Access Paper or Ask Questions

Private Hypothesis Selection


May 30, 2019
Mark Bun, Gautam Kamath, Thomas Steinke, Zhiwei Steven Wu


  Access Paper or Ask Questions

A Hybrid Approach to Privacy-Preserving Federated Learning


Dec 07, 2018
Stacey Truex, Nathalie Baracaldo, Ali Anwar, Thomas Steinke, Heiko Ludwig, Rui Zhang


  Access Paper or Ask Questions

The Limits of Post-Selection Generalization


Jun 15, 2018
Kobbi Nissim, Adam Smith, Thomas Steinke, Uri Stemmer, Jonathan Ullman


  Access Paper or Ask Questions

Calibrating Noise to Variance in Adaptive Data Analysis


Jun 11, 2018
Vitaly Feldman, Thomas Steinke

* Accepted for presentation at Conference on Learning Theory (COLT) 2018 

  Access Paper or Ask Questions

Generalization for Adaptively-chosen Estimators via Stable Median


Jun 15, 2017
Vitaly Feldman, Thomas Steinke

* To appear in Conference on Learning Theory (COLT) 2017 

  Access Paper or Ask Questions

Concentrated Differential Privacy: Simplifications, Extensions, and Lower Bounds


May 06, 2016
Mark Bun, Thomas Steinke


  Access Paper or Ask Questions

Make Up Your Mind: The Price of Online Queries in Differential Privacy


Apr 15, 2016
Mark Bun, Thomas Steinke, Jonathan Ullman


  Access Paper or Ask Questions

More General Queries and Less Generalization Error in Adaptive Data Analysis


Nov 10, 2015
Raef Bassily, Adam Smith, Thomas Steinke, Jonathan Ullman

* This paper was merged with another manuscript and is now subsumed by arXiv:1511.02513 

  Access Paper or Ask Questions

Algorithmic Stability for Adaptive Data Analysis


Nov 08, 2015
Raef Bassily, Kobbi Nissim, Adam Smith, Thomas Steinke, Uri Stemmer, Jonathan Ullman

* This work unifies and subsumes the two arXiv manuscripts arXiv:1503.04843 and arXiv:1504.05800 

  Access Paper or Ask Questions

Interactive Fingerprinting Codes and the Hardness of Preventing False Discovery


Feb 20, 2015
Thomas Steinke, Jonathan Ullman


  Access Paper or Ask Questions

Between Pure and Approximate Differential Privacy


Jan 24, 2015
Thomas Steinke, Jonathan Ullman


  Access Paper or Ask Questions

Weighted Polynomial Approximations: Limits for Learning and Pseudorandomness


Dec 08, 2014
Mark Bun, Thomas Steinke

* 22 pages 

  Access Paper or Ask Questions