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

As language models grow in popularity, it becomes increasingly important toclearly measure all possible markers of demographic identity in order to avoidperpetuating existing societal harms. Many datasets for measuring biascurrently exist, but they are restricted in their coverage of demographic axesand are commonly used with preset bias tests that presuppose which types ofbiases models can exhibit. In this work, we present a new, more inclusive biasmeasurement dataset, HolisticBias, which includes nearly 600 descriptor termsacross 13 different demographic axes. HolisticBias was assembled in aparticipatory process including experts and community members with livedexperience of these terms. These descriptors combine with a set of biasmeasurement templates to produce over 450,000 unique sentence prompts, which weuse to explore, identify, and reduce novel forms of bias in several generativemodels. We demonstrate that HolisticBias is effective at measuring previouslyundetectable biases in token likelihoods from language models, as well as in anoffensiveness classifier. We will invite additions and amendments to thedataset, which we hope will serve as a basis for more easy-to-use andstandardized methods for evaluating bias in NLP models.

"I'm sorry to hear that": Finding New Biases in Language Models with a Holistic Descriptor Dataset