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 BERT (mBERT) has demonstrated considerable cross-lingualsyntactic ability, whereby it enables effective zero-shot cross-lingualtransfer of syntactic knowledge. The transfer is more successful between somelanguages, but it is not well understood what leads to this variation andwhether it fairly reflects difference between languages. In this work, weinvestigate the distributions of grammatical relations induced from mBERT inthe context of 24 typologically different languages. We demonstrate that thedistance between the distributions of different languages is highly consistentwith the syntactic difference in terms of linguistic formalisms. Suchdifference learnt via self-supervision plays a crucial role in the zero-shottransfer performance and can be predicted by variation in morphosyntacticproperties between languages. These results suggest that mBERT properly encodeslanguages in a way consistent with linguistic diversity and provide insightsinto the mechanism of cross-lingual transfer.

Cross-Linguistic Syntactic Difference in Multilingual BERT: How Good is It and How Does It Affect Transfer?