English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Multilingual machine translation (MMT), trained on a mixture of parallel andmonolingual data, is key for improving translation in low-resource languagepairs. However, the literature offers conflicting results on the performance ofdifferent methods of including monolingual data. To resolve this, we examinehow denoising autoencoding (DAE) and backtranslation (BT) impact MMT underdifferent data conditions and model scales. Unlike prior studies, we use arealistic dataset of 100 translation directions and consider many domaincombinations of monolingual and test data. We find that monolingual datagenerally helps MMT, but models are surprisingly brittle to domain mismatches,especially at smaller model scales. BT is beneficial when the parallel,monolingual, and test data sources are similar but can be detrimentalotherwise, while DAE is less effective than previously reported. Next, weanalyze the impact of scale (from 90M to 1.6B parameters) and find it isimportant for both methods, particularly DAE. As scale increases, DAEtransitions from underperforming the parallel-only baseline at 90M toconverging with BT performance at 1.6B, and even surpassing it in low-resource.These results offer new insights into how to best use monolingual data in MMT.

When Does Monolingual Data Help Multilingual Translation: The Role of Domain and Model Scale