Performance in a dialectal profiling task of LLMs for varieties of Brazilian Portuguese
arXiv:2410.10991 · doi:10.5753/stil.2024.241891
Abstract
Different of biases are reproduced in LLM-generated responses, including dialectal biases. A study based on prompt engineering was carried out to uncover how LLMs discriminate varieties of Brazilian Portuguese, specifically if sociolinguistic rules are taken into account in four LLMs: GPT 3.5, GPT-4o, Gemini, and Sabi.-2. The results offer sociolinguistic contributions for an equity fluent NLP technology.
8 pages, XI Jornada de Descrição do Português