KT
Kunal Talwar
cs.LGcs.CRstat.MLcs.DSmath.OCmath.STstat.THmath.PRcs.CCcs.CG
On Valency
published · living versionsW_snpkpxfk·v1 · currentpublished
TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
with Martín Abadi, Ashish Agarwal, Paul Barham, Eugene Brevdo +35
1 version
Preprints & journals
82 papers in the corpus · 2007–2026Wally: Batched Private Nearest Neighbor Search at Scale2406.06761v7 · Hilal Asi, Fabian Boemer, Nicholas Genise et al.2024 · 0 citationsarXiv
Cram Less to Fit More: Training Data Pruning Improves Memorization of Facts2604.08519v1 · Jiayuan Ye, Vitaly Feldman, Kunal Talwar2026 · 0 citationsarXiv
The Importance of Being Smoothly Calibrated2603.16015v1 · Parikshit Gopalan, Konstantinos Stavropoulos, Kunal Talwar et al.2026 · 0 citationsarXiv
Enabling Differentially Private Federated Learning for Speech Recognition: Benchmarks, Adaptive Optimizers and Gradient Clipping2310.00098v4 · Martin Pelikan, Sheikh Shams Azam, Vitaly Feldman et al.2023 · 6 citationsarXiv
Efficient Calibration for Decision Making2511.13699v1 · Parikshit Gopalan, Konstantinos Stavropoulos, Kunal Talwar et al.2025 · 0 citationsarXiv
PREAMBLE: Private and Efficient Aggregation via Block Sparse Vectors2503.11897v2 · Hilal Asi, Vitaly Feldman, Hannah Keller et al.2025 · 0 citationsarXiv
Improved Sample Complexity for Private Nonsmooth Nonconvex Optimization2410.05880v2 · Guy Kornowski, Daogao Liu, Kunal Talwar2024 · 0 citationsarXiv
Instance-Optimality for Private KL Distribution Estimation2505.23620v1 · Jiayuan Ye, Vitaly Feldman, Kunal Talwar2025 · 0 citationsarXiv
Faster Rates for Private Adversarial Bandits2505.21790v1 · Hilal Asi, Vinod Raman, Kunal Talwar2025 · 0 citationsarXiv
On Privately Estimating a Single Parameter2503.17252v1 · Hilal Asi, John C. Duchi, Kunal Talwar2025 · 0 citationsarXiv
Adaptive Batch Size for Privately Finding Second-Order Stationary Points2410.07502v2 · Daogao Liu, Kunal Talwar2024 · 0 citationsarXiv
Private Online Learning via Lazy Algorithms2406.03620v2 · Hilal Asi, Tomer Koren, Daogao Liu et al.2024 · 1 citationarXiv
Fingerprinting Codes Meet Geometry: Improved Lower Bounds for Private Query Release and Adaptive Data Analysis2412.14396v1 · Xin Lyu, Kunal Talwar2024 · 1 citationarXiv
Privacy-Computation trade-offs in Private Repetition and Metaselection2410.19012v1 · Kunal Talwar2024 · 0 citationsarXiv
Concentration of the Langevin Algorithm's Stationary Distribution2212.12629v2 · Jason M. Altschuler, Kunal Talwar2022 · 0 citationsarXiv
Samplable Anonymous Aggregation for Private Federated Data Analysis2307.15017v2 · Kunal Talwar, Shan Wang, Audra McMillan et al.2023 · 10 citationsarXiv
Instance-Optimal Private Density Estimation in the Wasserstein Distance2406.19566v1 · Vitaly Feldman, Audra McMillan, Satchit Sivakumar et al.2024 · 1 citationarXiv
PINE: Efficient Norm-Bound Verification for Secret-Shared Vectors2311.10237v2 · Guy N. Rothblum, Eran Omri, Junye Chen et al.2023 · 0 citationsarXiv
Private Vector Mean Estimation in the Shuffle Model: Optimal Rates Require Many Messages2404.10201v2 · Hilal Asi, Vitaly Feldman, Jelani Nelson et al.2024 · 1 citationarXiv
Stronger Privacy Amplification by Shuffling for R\'enyi and Approximate Differential Privacy2208.04591v2 · Vitaly Feldman, Audra McMillan, Kunal Talwar2022 · 0 citationsarXiv
Mean Estimation with User-level Privacy under Data Heterogeneity2307.15835v1 · Rachel Cummings, Vitaly Feldman, Audra McMillan et al.2023 · 7 citationsarXiv
Differentially Private Heavy Hitter Detection using Federated Analytics2307.11749v1 · Karan Chadha, Junye Chen, John Duchi et al.2023 · 9 citationsarXiv
Private Online Prediction from Experts: Separations and Faster Rates2210.13537v3 · Hilal Asi, Vitaly Feldman, Tomer Koren et al.2022 · 0 citationsarXiv
Fast Optimal Locally Private Mean Estimation via Random Projections2306.04444v2 · Hilal Asi, Vitaly Feldman, Jelani Nelson et al.2023 · 3 citationsarXiv
Privacy of Noisy Stochastic Gradient Descent: More Iterations without More Privacy Loss2205.13710v2 · Jason M. Altschuler, Kunal Talwar2022 · 13 citationsarXiv
Near-Optimal Algorithms for Private Online Optimization in the Realizable Regime2302.14154v1 · Hilal Asi, Vitaly Feldman, Tomer Koren et al.2023 · 0 citationsarXiv
Private Federated Statistics in an Interactive Setting2211.10082v1 · Audra McMillan, Omid Javidbakht, Kunal Talwar et al.2022 · 3 citationsarXiv
Resolving the Mixing Time of the Langevin Algorithm to its Stationary Distribution for Log-Concave Sampling2210.08448v2 · Jason M. Altschuler, Kunal Talwar2022 · 3 citationsarXiv
Subspace Recovery from Heterogeneous Data with Non-isotropic Noise2210.13497v1 · John Duchi, Vitaly Feldman, Lunjia Hu et al.2022 · 1 citationarXiv
FLAIR: Federated Learning Annotated Image Repository2207.08869v1 · Congzheng Song, Filip Granqvist, Kunal Talwar2022 · 7 citationsarXiv
Optimal Algorithms for Mean Estimation under Local Differential Privacy2205.02466v1 · Hilal Asi, Vitaly Feldman, Kunal Talwar2022 · 4 citationsarXiv
Private Frequency Estimation via Projective Geometry2203.00194v1 · Vitaly Feldman, Jelani Nelson, Huy Le Nguyen et al.2022 · 2 citationsarXiv
Differential Secrecy for Distributed Data and Applications to Robust Differentially Secure Vector Summation2202.10618v1 · Kunal Talwar2022 · 82 citationsarXiv
Online learning over a finite action set with limited switching1803.01548v2 · Jason Altschuler, Kunal Talwar2018 · 7 citationsMathematics of Operations Research, 46(1), 179-203, 2021
Hiding Among the Clones: A Simple and Nearly Optimal Analysis of Privacy Amplification by Shuffling2012.12803v3 · Vitaly Feldman, Audra McMillan, Kunal Talwar2020 · 84 citationsarXiv
Private Adaptive Gradient Methods for Convex Optimization2106.13756v1 · Hilal Asi, John Duchi, Alireza Fallah et al.2021 · 15 citationsarXiv
Characterizing Structural Regularities of Labeled Data in Overparameterized Models2002.03206v3 · Ziheng Jiang, Chiyuan Zhang, Kunal Talwar et al.2020 · 28 citationsarXiv
Private Stochastic Convex Optimization: Optimal Rates in $\ell_1$ Geometry2103.01516v1 · Hilal Asi, Vitaly Feldman, Tomer Koren et al.2021 · 4 citationsarXiv
Lossless Compression of Efficient Private Local Randomizers2102.12099v1 · Vitaly Feldman, Kunal Talwar2021 · 7 citationsarXiv
Cops, Robbers, and Threatening Skeletons: Padded Decomposition for Minor-Free Graphs1311.3048v2 · Ittai Abraham, Cyril Gavoille, Anupam Gupta et al.2013 · 28 citationsarXiv
Faster Differentially Private Samplers via R\'enyi Divergence Analysis of Discretized Langevin MCMC2010.14658v2 · Arun Ganesh, Kunal Talwar2020 · 4 citationsarXiv
On the Error Resistance of Hinge Loss Minimization2012.00989v1 · Kunal Talwar2020 · 0 citationsarXiv
Stochastic Optimization with Laggard Data Pipelines2010.13639v1 · Naman Agarwal, Rohan Anil, Tomer Koren et al.2020 · 4 citationsarXiv
Amplification by Shuffling: From Local to Central Differential Privacy via Anonymity1811.12469v2 · 'Ulfar Erlingsson, Vitaly Feldman, Ilya Mironov et al.2018 · 278 citationsarXiv
Stability of Stochastic Gradient Descent on Nonsmooth Convex Losses2006.06914v1 · Raef Bassily, Vitaly Feldman, Crist'obal Guzm'an et al.2020 · 18 citationsarXiv
Private Stochastic Convex Optimization: Optimal Rates in Linear Time2005.04763v1 · Vitaly Feldman, Tomer Koren, Kunal Talwar2020 · 77 citationsarXiv
Encode, Shuffle, Analyze Privacy Revisited: Formalizations and Empirical Evaluation2001.03618v1 · 'Ulfar Erlingsson, Vitaly Feldman, Ilya Mironov et al.2020 · 35 citationsarXiv
Computational Separations between Sampling and Optimization1911.02074v1 · Kunal Talwar2019 · 5 citationsarXiv
R\'enyi Differential Privacy of the Sampled Gaussian Mechanism1908.10530v1 · Ilya Mironov, Kunal Talwar, Li Zhang2019 · 83 citationsarXiv
Private Stochastic Convex Optimization with Optimal Rates1908.09970v1 · Raef Bassily, Vitaly Feldman, Kunal Talwar et al.2019 · 81 citationsarXiv
Career total: 214 works. 82 are in this corpus.Showing the 50 most recent.
Profile built from the corpus for this byline.
Author records are still filling in while the Hub is in alpha. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.