
다양한 언어 및 지역의 사회적 편견을 반영하는 SeeGULL Multilingual 다국어 스테레오타입 데이터셋을 구축하고 모델 평가에 활용한다.


coremsg

seegull-multilingual-a-multilingual-dataset-of-stereotypes


SeeGULL Multilingual: A Multilingual Dataset of Stereotypes


title_rewrite


Generative multilingual models lack cross-cultural considerations in safety evaluations, necessitating a global-scale stereotype dataset like SeeGULL Multilingual.


seegull-multilingual-a-global-dataset-of-social-stereotypes-across-20-languages-and-23-regions


SeeGULL Multilingual: A Global Dataset of Social Stereotypes Across 20 Languages and 23 Regions



The author introduces SeeGULL Multilingual, a dataset of global stereotypes, to address the lack of cross-cultural considerations in generative multilingual models' safety and fairness evaluations.


seegull-multilingual-a-dataset-of-global-stereotypes


SeeGULL Multilingual: A Dataset of Global Stereotypes
