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

The language ability of Large Language Models (LLMs) is often unbalancedtowards English because of the imbalance in the distribution of thepre-training data. This disparity is demanded in further fine-tuning andaffecting the cross-lingual abilities of LLMs. In this paper, we propose toempower Instructiontuned LLMs (It-LLMs) in languages other than English bybuilding semantic alignment between them. Hence, we propose CrossAlpaca, anIt-LLM with cross-lingual instruction-following and Translation-followingdemonstrations to improve semantic alignment between languages. We validate ourapproach on the multilingual Question Answering (QA) benchmarks XQUAD and MLQAand adapted versions of MMLU and BBH. Our models, tested over six differentlanguages, outperform the It-LLMs tuned on monolingual data. The final resultsshow that instruction tuning on non-English data is not enough and thatsemantic alignment can be further improved by Translation-followingdemonstrations.

Empowering Cross-lingual Abilities of Instruction-tuned Large Language Models by Translation-following demonstrations