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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Multilingual Neural Machine Translation (MNMT) facilitates knowledge sharingbut often suffers from poor zero-shot (ZS) translation qualities. While priorwork has explored the causes of overall low ZS performance, our work introducesa fresh perspective: the presence of high variations in ZS performance. Thissuggests that MNMT does not uniformly exhibit poor ZS capability; instead,certain translation directions yield reasonable results. Through systematicexperimentation involving 1,560 language directions spanning 40 languages, weidentify three key factors contributing to high variations in ZS NMTperformance: 1) target side translation capability 2) vocabulary overlap 3)linguistic properties. Our findings highlight that the target side translationquality is the most influential factor, with vocabulary overlap consistentlyimpacting ZS performance. Additionally, linguistic properties, such as languagefamily and writing system, play a role, particularly with smaller models.Furthermore, we suggest that the off-target issue is a symptom of inadequate ZSperformance, emphasizing that zero-shot translation challenges extend beyondaddressing the off-target problem. We release the data and models serving as abenchmark to study zero-shot for future research athttps://github.com/Smu-Tan/ZS-NMT-Variations

Towards a Better Understanding of Variations in Zero-Shot Neural Machine Translation Performance