Marked Personas: Using Natural Language Prompts to Measure Stereotypes in Language Models
Authors:
Preprint
Abstract
To recognize and mitigate harms from large language models (LLMs), we need to understand the prevalence and nuances of stereotypes in LLM outputs. Toward this end, we present Marked Personas, a prompt-based method to measure stereotypes in LLMs for intersectional demographic groups without any lexicon or data labeling. Grounded in the sociolinguistic concept of markedness (which characterizes explicitly linguistically marked categories versus unmarked defaults), our proposed method is twofold: 1) prompting an LLM to generate personas, i.e., natural language descriptions, of the target demographic group alongside personas of unmarked, default groups; 2) identifying the words that significantly distinguish personas of the target group from corresponding unmarked ones. We find that the portrayals generated by GPT-3.5 and GPT-4 contain higher rates of racial stereotypes than human-written portrayals using the same prompts. The words distinguishing personas of marked (non-white, non-male) groups reflect patterns of othering and exoticizing these demographics. An intersectional lens further reveals tropes that dominate portrayals of marginalized groups, such as tropicalism and the hypersexualization of minoritized women. These representational harms have concerning implications for downstream applications like story generation.Keywords
All Related Versions
This publication has 28 references indexed in Scilit:
- Misogynistic terrorism: it has always been hereCritical Studies on Terrorism, 2022
- Creating and Undoing Legacies of Resilience: Black Women as Martyrs in the Black Community Under Oppressive Social ControlWomen & Criminal Justice, 2020
- Gender stereotypes have changed: A cross-temporal meta-analysis of U.S. public opinion polls from 1946 to 2018.American Psychologist, 2020
- The Rise of Neoliberal FeminismCultural Studies, 2013
- Awakenings: The Emergence of White Supremacist IdeologiesDeviant Behavior, 2013
- Common racial stereotypesEqual Opportunities International, 2003
- Description and Prescription: How Gender Stereotypes Prevent Women's Ascent Up the Organizational LadderJournal of Social Issues, 2001
- The Inmates are Running the Asylum1999
- The Orientalization of Asian women in AmericaWomen's Studies International Forum, 1998
- “Holy terror”: The implications of terrorism motivated by a religious imperativeStudies in Conflict & Terrorism, 1995