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 models have demonstrated impressive cross-lingual transferperformance. However, test sets like XNLI are monolingual at the example level.In multilingual communities, it is common for polyglots to code-mix whenconversing with each other. Inspired by this phenomenon, we present two strongblack-box adversarial attacks (one word-level, one phrase-level) formultilingual models that push their ability to handle code-mixed sentences tothe limit. The former uses bilingual dictionaries to propose perturbations andtranslations of the clean example for sense disambiguation. The latter directlyaligns the clean example with its translations before extracting phrases asperturbations. Our phrase-level attack has a success rate of 89.75% againstXLM-R-large, bringing its average accuracy of 79.85 down to 8.18 on XNLI.Finally, we propose an efficient adversarial training scheme that trains in thesame number of steps as the original model and show that it improves modelaccuracy.

Code-Mixing on Sesame Street: Dawn of the Adversarial Polyglots