Visualizing LSTM Networks. Part I. – Acta Schola Automata Polonica – Medium

Long-Short Term Memory networks are state-of-the-art tools for long sequence modeling. However, there is a problem with understanding what they have learned and investigating why they are making particular mistakes. Many articles and papers do it for convolutional neural networks, but for LSTM we do not have many tools to look inside and debug them. In this article we try to partially fill this gap. We visualize LSTM network activations from Australian sign language (Auslan) sign classifying model.