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

Modern multilingual models are trained on concatenated text from multiplelanguages in hopes of conferring benefits to each (positive transfer), with themost pronounced benefits accruing to low-resource languages. However, recentwork has shown that this approach can degrade performance on high-resourcelanguages, a phenomenon known as negative interference. In this paper, wepresent the first systematic study of negative interference. We show that,contrary to previous belief, negative interference also impacts low-resourcelanguages. While parameters are maximally shared to learn language-universalstructures, we demonstrate that language-specific parameters do exist inmultilingual models and they are a potential cause of negative interference.Motivated by these observations, we also present a meta-learning algorithm thatobtains better cross-lingual transferability and alleviates negativeinterference, by adding language-specific layers as meta-parameters andtraining them in a manner that explicitly improves shared layers'generalization on all languages. Overall, our results show that negativeinterference is more common than previously known, suggesting new directionsfor improving multilingual representations.

On Negative Interference in Multilingual Models: Findings and A Meta-Learning Treatment