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

Low-resource languages, such as Baltic languages, benefit from LargeMultilingual Models (LMs) that possess remarkable cross-lingual transferperformance capabilities. This work is an interpretation and analysis studyinto cross-lingual representations of Multilingual LMs. Previous workshypothesized that these LMs internally project representations of differentlanguages into a shared cross-lingual space. However, the literature producedcontradictory results. In this paper, we revisit the prior work claiming that"BERT is not an Interlingua" and show that different languages do converge to ashared space in such language models with another choice of pooling strategy orsimilarity index. Then, we perform cross-lingual representational analysis forthe two most popular multilingual LMs employing 378 pairwise languagecomparisons. We discover that while most languages share joint cross-lingualspace, some do not. However, we observe that Baltic languages do belong to thatshared space. The code is available at https://github.com/TartuNLP/xsim.

Similarity of Sentence Representations in Multilingual LMs: Resolving Conflicting Literature and Case Study of Baltic Languages