deep learning Interpretable linguistic representations learned by LSTMs Inspired by semantic representations in biology, I present interpretable representations learned by Long Short-Term Memory (LSTM) hidden state neurons, in an effort to explain their predictive success.
