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

Pretrained multilingual models enable zero-shot learning even for unseenlanguages, and that performance can be further improved via adaptation prior tofinetuning. However, it is unclear how the number of pretraining languagesinfluences a model's zero-shot learning for languages unseen duringpretraining. To fill this gap, we ask the following research questions: (1) Howdoes the number of pretraining languages influence zero-shot performance onunseen target languages? (2) Does the answer to that question change with modeladaptation? (3) Do the findings for our first question change if the languagesused for pretraining are all related? Our experiments on pretraining withrelated languages indicate that choosing a diverse set of languages is crucial.Without model adaptation, surprisingly, increasing the number of pretraininglanguages yields better results up to adding related languages, after whichperformance plateaus. In contrast, with model adaptation via continuedpretraining, pretraining on a larger number of languages often gives furtherimprovement, suggesting that model adaptation is crucial to exploit additionalpretraining languages.

Match the Script, Adapt if Multilingual: Analyzing the Effect of Multilingual Pretraining on Cross-lingual Transferability