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

To recognize and mitigate harms from large language models (LLMs), we need tounderstand the prevalence and nuances of stereotypes in LLM outputs. Towardthis end, we present Marked Personas, a prompt-based method to measurestereotypes in LLMs for intersectional demographic groups without any lexiconor data labeling. Grounded in the sociolinguistic concept of markedness (whichcharacterizes explicitly linguistically marked categories versus unmarkeddefaults), our proposed method is twofold: 1) prompting an LLM to generatepersonas, i.e., natural language descriptions, of the target demographic groupalongside personas of unmarked, default groups; 2) identifying the words thatsignificantly distinguish personas of the target group from correspondingunmarked ones. We find that the portrayals generated by GPT-3.5 and GPT-4contain higher rates of racial stereotypes than human-written portrayals usingthe same prompts. The words distinguishing personas of marked (non-white,non-male) groups reflect patterns of othering and exoticizing thesedemographics. An intersectional lens further reveals tropes that dominateportrayals of marginalized groups, such as tropicalism and thehypersexualization of minoritized women. These representational harms haveconcerning implications for downstream applications like story generation.

Marked Personas: Using Natural Language Prompts to Measure Stereotypes in Language Models