HESEIA: A community-based dataset for evaluating social biases in large language models, co-designed in real school settings in Latin America
This addresses the problem of evaluating social biases in LLMs for educational communities in Latin America, offering a more participatory and contextually relevant benchmark, though it is incremental in dataset creation.
The authors tackled the lack of community involvement in bias evaluation datasets by introducing HESEIA, a dataset of 46,499 sentences co-designed with 370 teachers and 5,370 students in Latin America, which captures more unrecognized stereotypes in LLMs than previous datasets.
Most resources for evaluating social biases in Large Language Models are developed without co-design from the communities affected by these biases, and rarely involve participatory approaches. We introduce HESEIA, a dataset of 46,499 sentences created in a professional development course. The course involved 370 high-school teachers and 5,370 students from 189 Latin-American schools. Unlike existing benchmarks, HESEIA captures intersectional biases across multiple demographic axes and school subjects. It reflects local contexts through the lived experience and pedagogical expertise of educators. Teachers used minimal pairs to create sentences that express stereotypes relevant to their school subjects and communities. We show the dataset diversity in term of demographic axes represented and also in terms of the knowledge areas included. We demonstrate that the dataset contains more stereotypes unrecognized by current LLMs than previous datasets. HESEIA is available to support bias assessments grounded in educational communities.