Posted Apr 10 2018
"Data as a mirror of society: lessons from the emerging science of bias in machine learning"
Thursday, April 12, 2018 - 4:30pm
A32 Princeton Neuroscience Institute
Ahmed El Hady and Aaron M. Bornstein
PNI In-house Seminar Series
Arvind Narayanan is an Assistant Professor in the Department of Computer Science. He will talk about research uncovering latent social prejudices in language models. He and colleagues developed a machine learning version of the Implicit Association Test (IAT). Applied to standard word embedding databases, the test recovers a wide range of prejudices documented in human behavior. This work raises questions about how prejudice is encoded and how it is transmitted, with potential implications for social neuroscience, linguistics, and the correspondence between representations in artificial and biological neural networks.