SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2101.01761 · AAAI · 2021

AutoDropout: Learning Dropout Patterns to Regularize Deep Networks

V Le, Hieu Pham And Quoc

arXiv · PDF · Open in the Atlas

Code that ran

We have not lifted any functions out of this paper's repositories yet, so there is nothing we have run. The repositories linked to it are listed below.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Neural networks are often over-parameterized and hence benefit from aggressive regularization. Conventional regularization methods, such as Dropout or weight decay, do not leverage the structures of the network's inputs and hidden states. As a result, these methods are less effective than recent methods that leverage the structures, such as SpatialDropout and Drop-Block , which randomly drop the values at certain contiguous areas in the hidden states and setting them to zero. Although the locations of dropout areas are random, the patterns of SpatialDropout and DropBlock are manually designed and fixed. Here we propose AutoDropout, which automates the process of designing dropout patterns. In our method, a controller learns to generate a dropout pattern at every channel and layer of a target network, such as a ConvNet or a Transformer. The target network is then trained with the dropout pattern, and its resulting validation performance is used as a signal for the controller to learn from. We show that this method works well for both image recognition on CIFAR-10 and Ima-geNet, as well as language modeling on Penn Treebank and WikiText-2. The learned dropout patterns also transfers to different tasks and datasets, such as from language model on Penn Treebank to Engligh-French translation on WMT 2014. Our code will be available at: https://github.com/googleresearch/google-research/tree/master/auto_dropout.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2101.01761")
get_code_for_paper("2101.01761")
have("2101.01761")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2101.01761.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2101.01761.svg)](https://syntology.ai/paper/2101.01761)

Connect an agent — have() is free.