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 social biases and unwelcome stereotypes revealed by pretrained languagemodels are becoming obstacles to their application. Compared to numerousdebiasing methods targeting word level, there has been relatively lessattention on biases present at phrase level, limiting the performance ofdebiasing in discipline domains. In this paper, we propose an automaticmulti-token debiasing pipeline called \textbf{General Phrase Debiaser}, whichis capable of mitigating phrase-level biases in masked language models.Specifically, our method consists of a \textit{phrase filter stage} thatgenerates stereotypical phrases from Wikipedia pages as well as a \textit{modeldebias stage} that can debias models at the multi-token level to tackle biaschallenges on phrases. The latter searches for prompts that trigger model'sbias, and then uses them for debiasing. State-of-the-art results on standarddatasets and metrics show that our approach can significantly reduce genderbiases on both career and multiple disciplines, across models with varyingparameter sizes.

General Phrase Debiaser: Debiasing Masked Language Models at a Multi-Token Level