Methods › General › Regularization › Label Smoothing

Label Smoothing

introduced 1985 14,327 papers tagged archive 2025-07-28

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

Label Smoothing is a regularization technique that introduces noise for the labels. This accounts for the fact that datasets may have mistakes in them, so maximizing the likelihood of logp(y|x) directly can be harmful. Assume for a small constant ϵ, the training set label y is correct with probability 1-ϵ and incorrect otherwise. Label Smoothing regularizes a model based on a softmax with k output values by replacing the hard $0$ and $1$ classification targets with targets of ϵ/k and 1-(k-1)/kϵ respectively.

Source: Deep Learning, Goodfellow et al

Image Source: When Does Label Smoothing Help?

Papers archive 2025-07-28

30 shown of 14,327, 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 2,170 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 Modelling1,245
Decoder1,065
Language Modeling955
Translation777
Machine Translation739
Semantic Segmentation688
Object Detection612
Image Classification575
object-detection553
Question Answering514
Retrieval460
Sentence457
Segmentation454
image-classification453
Representation Learning423
Large Language Model409
Object379
Time Series375
Transfer Learning323
Text Generation317

Usage over time archive 2025-07-28

Papers per year tagged with Label Smoothing: 2015 to 2025, peak 4,388 4,388 0 2015: 1 paper 2015 2016: 4 papers 2016 2017: 30 papers 2017 2018: 131 papers 2018 2019: 540 papers 2019 2020: 901 papers 2020 2021: 1534 papers 2021 2022: 1934 papers 2022 2023: 3267 papers 2023 2024: 4388 papers 2024 2025: 1597 papers 2025
Papers per year the archive tags with this method, by the paper's archive date (14,327 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