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

LLMs, Personas