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

Human perception of language depends on personal backgrounds like gender andethnicity. While existing studies have shown that large language models (LLMs)hold values that are closer to certain societal groups, it is unclear whethertheir prediction behaviors on subjective NLP tasks also exhibit a similar bias.In this study, leveraging the POPQUORN dataset which contains annotations ofdiverse demographic backgrounds, we conduct a series of experiments on fourpopular LLMs to investigate their capability to understand group differencesand potential biases in their predictions for politeness and offensiveness. Wefind that for both tasks, model predictions are closer to the labels from Whiteand female participants. We further explore prompting with the targetdemographic labels and show that including the target demographic in the promptactually worsens the model's performance. More specifically, when beingprompted to respond from the perspective of "Black" and "Asian" individuals,models show lower performance in predicting both overall scores as well as thescores from corresponding groups. Our results suggest that LLMs hold gender andracial biases for subjective NLP tasks and that demographic-infused promptsalone may be insufficient to mitigate such effects. Code and data are availableat https://github.com/Jiaxin-Pei/LLM-Group-Bias.

Aligning with Whom? Large Language Models Have Gender and Racial Biases in Subjective NLP Tasks