Studying biases in word embeddings
Gender de-biasing of word embeddings, via neutralization and equalization algorithms.
Based on an assignment from the Deep Learning course Sequential Models offered by deeplearning.ai, and the Bolukbasi et al paper from 2016 (https://arxiv.org/pdf/1607.06520.pdf).