MIT’s new explainability technique can auto- describe the roles of individual neurons in a neural network in an NLU February 7, 2022 // by Finnovate This content is for members only. Sign up for access to the latest trends and innovations in fintech. View subscription plans. Login