1 paper
Peter W. Chang, Leor Fishman, Seth Neel
It is widely held that one cause of downstream bias in classifiers is bias present in the training data. Rectifying such biases may involve context-dependent interventions such as…