Methods › General › Regularization › Variational Dropout

Variational Dropout

141 papers tagged archive 2025-07-28

Introduced by Yarin Gal et al. in A Theoretically Grounded Application of Dropout in Recurrent Neural Networks

archive 2025-07-28 Description, source and code snippet are the archive's method entry.

Variational Dropout is a regularization technique based on dropout, but uses a variational inference grounded approach. In Variational Dropout, we repeat the same dropout mask at each time step for both inputs, outputs, and recurrent layers (drop the same network units at each time step). This is in contrast to ordinary Dropout where different dropout masks are sampled at each time step for the inputs and outputs alone.

PaperSourceSee Code · salesforce/awd-lstm-lm

Papers archive 2025-07-28

30 shown of 141, newest first. Repository counts are the archive's code-links table. A Syntology line states what Syntology ran from that paper's harvested code; it is per sample and not a correctness claim.

Tasks archive 2025-07-28

20 shown of 144 tasks the archive attaches to papers tagged with this method, by distinct papers. A task without a page in the catalog is plain text.

TaskPapers
Language Modelling62
Language Modeling46
Text Classification17
Transfer Learning16
General Classification15
Sentiment Analysis13
text-classification12
Decoder9
Classification8
Machine Translation8
Translation8
Speech Recognition7
Sentence6
Word Embeddings6
speech-recognition6
Image Classification5
Variational Inference5
Automatic Speech Recognition4
Automatic Speech Recognition (ASR)4
Decision Making4

Usage over time archive 2025-07-28

Papers per year tagged with Variational Dropout: 2015 to 2025, peak 35 35 0 2015: 1 paper 2015 2016: 2 papers 2016 2017: 7 papers 2017 2018: 6 papers 2018 2019: 29 papers 2019 2020: 35 papers 2020 2021: 32 papers 2021 2022: 7 papers 2022 2023: 13 papers 2023 2024: 6 papers 2024 2025: 3 papers 2025
Papers per year the archive tags with this method, by the paper's archive date (141 dated). Bars are counts, not a trend claim.

Components: the archive holds no method-to-method composition, so PwC's Components table cannot be rebuilt; the Papers list carries no Results column for the same reason (the archive does not join its leaderboard rows to method tags).

Categories archive 2025-07-28

Regularization

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections