LATIDIA · Investigación

GRAP: Sondeo consciente de la rúbrica de género para analizar el sesgo de género en los LLM

arXiv: 2611.13024v1Tipo de anuncio: nuevo Resumen: A medida que los LLM median cada vez más en las interacciones cotidianas, los sutiles sesgos de género en las respuestas abiertas plantean riesgos sociales crecientes. Los métodos de sondeo existentes tienen dos limites clave

La noticia

arXiv:2611.13024v1 Announce Type: new Abstract: As LLMs increasingly mediate everyday interactions, subtle gender biases in open-ended responses pose growing societal risks. Existing probing methods have two key limitations: they rely on fixed-format tasks, such as pronoun resolution or occupation association, that alignment-tuned models can easily handle; and they compare raw group-level rubric scores, ignoring per-dimension variance and obscuring which bias dimensions reliably distinguish gender groups. We introduce GRAP (Gender Rubric-Aware Probing), a framework that uses a lightweight logistic probe to identify the most discriminative bias dimensions from responses scored across 10 rubric dimensions, operating task-agnostically on open-ended instruction following.