Using Adversarial Debiasing to Remove Bias from Word Embeddings

Dana Kenna

Word Embeddings have been shown to contain the societal biases present in the original corpora. Existing methods to deal with this problem have been shown to only remove superficia…