Tomasz Limisiewicz
@tomlimi
CL Ph.D. student @CharlesUniPRG Interested in going into the inner workings of neural networks, multilingualism, and fairer NLP (he/him)
ID: 1436653954289606657
11-09-2021 11:33:08
173 Tweet
350 Followers
449 Following
Our new paper will appear at genderbiasnlp. We consider two types of gender signals in contextual embeddings of LMs: factual gender (e.g., denoted in the pronouns) and gender bias (stereotypical association between a word and a specific gender, e.g., doctor and nurse)