(Translated by https://www.hiragana.jp/)
[2406.00020] Harmful Speech Detection by Language Models Exhibits Gender-Queer Dialect Bias