Assessing Social and Intersectional Biases in Contextualized Word Representations

#artificialintelligence 

[no summary]