activity
20152021
most citedAttack of the Tails: Yes, You Really Can Backdoor Federated Learning

110 citations · 147 across the 5 of their papers we have counts for

collaborators

8 papers

cs.LG202116 cited

AutoFreeze: Automatically Freezing Model Blocks to Accelerate Fine-tuning

Yuhan Liu, Saurabh Agarwal, Shivaram Venkataraman

With the rapid adoption of machine learning (ML), a number of domains now use the approach of fine tuning models which were pre-trained on a large corpus of data. However, our expe…

cs.LG20215 cited

Pufferfish: Communication-efficient Models At No Extra Cost

Hongyi Wang, Saurabh Agarwal, Dimitris Papailiopoulos

To mitigate communication overheads in distributed model training, several studies propose the use of compressed stochastic gradients, usually achieved by sparsification or quantiz…

cs.DC2021

On the Utility of Gradient Compression in Distributed Training Systems

Saurabh Agarwal, Hongyi Wang, Shivaram Venkataraman +1

A rich body of prior work has highlighted the existence of communication bottlenecks in synchronous data-parallel training. To alleviate these bottlenecks, a long line of recent wo…

cs.LG202015 cited

Accordion: Adaptive Gradient Communication via Critical Learning Regime Identification

Saurabh Agarwal, Hongyi Wang, Kangwook Lee +2

Distributed model training suffers from communication bottlenecks due to frequent model updates transmitted across compute nodes. To alleviate these bottlenecks, practitioners use…

cs.LG2020110 cited

Attack of the Tails: Yes, You Really Can Backdoor Federated Learning

Hongyi Wang, Kartik Sreenivasan, Shashank Rajput +5

Due to its decentralized nature, Federated Learning (FL) lends itself to adversarial attacks in the form of backdoors during training. The goal of a backdoor is to corrupt the perf…

cs.LG2019

Scalable K-Medoids via True Error Bound and Familywise Bandits

Aravindakshan Babu, Saurabh Agarwal, Sudarshan Babu +1

K-Medoids(KM) is a standard clustering method, used extensively on semi-metric data.Error analyses of KM have traditionally used an in-sample notion of error,which can be far from…