Social Bias Evaluation for Large Language Models Requires Prompt Variations.

Rem Hida, Masahiro Kaneko, Naoaki Okazaki

Browse the full EMNLP paper archive.