activity
20192022
most citedTime Matters in Regularizing Deep Networks: Weight Decay and Data Augmentation Affect Early Learning Dynamics, Matter Little Near Convergence

21 citations · 21 across the 2 of their papers we have counts for

collaborators

6 papers

cs.CV2022

Mixed Differential Privacy in Computer Vision

Aditya Golatkar, Alessandro Achille, Yu-Xiang Wang +3

We introduce AdaMix, an adaptive differentially private algorithm for training deep neural network classifiers using both private and public image data. While pre-training language…

cs.LG2020

LQF: Linear Quadratic Fine-Tuning

Alessandro Achille, Aditya Golatkar, Avinash Ravichandran +2

Classifiers that are linear in their parameters, and trained by optimizing a convex loss function, have predictable behavior with respect to changes in the training data, initial c…

cs.LG2020

Mixed-Privacy Forgetting in Deep Networks

Aditya Golatkar, Alessandro Achille, Avinash Ravichandran +2

We show that the influence of a subset of the training samples can be removed -- or "forgotten" -- from the weights of a network trained on large-scale image classification tasks,…

cs.LG2020

Forgetting Outside the Box: Scrubbing Deep Networks of Information Accessible from Input-Output Observations

Aditya Golatkar, Alessandro Achille, Stefano Soatto

We describe a procedure for removing dependency on a cohort of training data from a trained deep network that improves upon and generalizes previous methods to different readout fu…

cs.LG2019

Eternal Sunshine of the Spotless Net: Selective Forgetting in Deep Networks

Aditya Golatkar, Alessandro Achille, Stefano Soatto

We explore the problem of selectively forgetting a particular subset of the data used for training a deep neural network. While the effects of the data to be forgotten can be hidde…

cs.LG201921 cited

Time Matters in Regularizing Deep Networks: Weight Decay and Data Augmentation Affect Early Learning Dynamics, Matter Little Near Convergence

Aditya Golatkar, Alessandro Achille, Stefano Soatto

Regularization is typically understood as improving generalization by altering the landscape of local extrema to which the model eventually converges. Deep neural networks (DNNs),…