Understanding Down Syndrome Stereotypes in LLM-Based Personas

TL;DR

This paper investigates how large language models can unintentionally generate stereotypes of people with Down syndrome, emphasizing the importance of participatory design to better capture their diverse experiences.

Contribution

It introduces stereotype detection in LLM-based personas and analyzes how data, interface, and tone influence stereotype emergence.

Findings

Stereotypes can emerge from training data and interface design.

Caregiver interviews reveal challenges in defining stereotypes.

Participatory methods are needed to represent heterogeneity.

Abstract

We present a case study of Persona-L, a system that leverages large language models (LLMs) and retrieval-augmented generation (RAG) to model personas of people with Down syndrome. Existing approaches to persona creation can often lead to oversimplified or stereotypical profiles of people with Down Syndrome. To that end, we built stereotype detection capabilities into Persona-L. Through interviews with caregivers and healthcare professionals (N=10), we examine how Down Syndrome stereotypes could manifest in both, content and delivery of LLMs, and interface design. Our findings show the challenges in stereotypes definition, and reveal the potential stereotype emergence from the training data, interface design, and the tone of LLM output. This highlights the need for participatory methods that capture the heterogeneity of lived experiences of people with Down Syndrome.