Understanding Hidden Memories of Recurrent Neural Networks
Authors:
- 1 October 2017
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 13-24
Abstract
Recurrent neural networks (RNNs) have been successfully applied to various natural language processing (NLP) tasks and achieved better results than conventional methods. However, the lack of understanding of the mechanisms behind their effectiveness limits further improvements on their architectures. In this paper, we present a visual analytics method for understanding and comparing RNN models for NLP tasks. We propose a technique to explain the function of individual hidden state units based on their expected response to input texts. We then co-cluster hidden state units and words based on the expected response and visualize co-clustering results as memory chips and word clouds to provide more structured knowledge on RNNs' hidden states. We also propose a glyph-based sequence visualization based on aggregate information to analyze the behavior of an RNN's hidden state at the sentence-level. The usability and effectiveness of our method are demonstrated through case studies and reviews from domain experts.Keywords
All Related Versions
This publication has 28 references indexed in Scilit:
- Interactive visual co-cluster analysis of bipartite graphs2016
- Document Modeling with Gated Recurrent Neural Network for Sentiment Classification2015
- A Five-Level Design Framework for Bicluster VisualizationsIEEE Transactions on Visualization and Computer Graphics, 2014
- ConVis: A Visual Text Analytic System for Exploring Blog ConversationsComputer Graphics Forum, 2014
- On the Properties of Neural Machine Translation: Encoder–Decoder Approaches2014
- Speech recognition with deep recurrent neural networks2013
- PivotPaths: Strolling through Faceted Information SpacesIEEE Transactions on Visualization and Computer Graphics, 2012
- Biclustering algorithms for biological data analysis: a surveyIEEE/ACM Transactions on Computational Biology and Bioinformatics, 2004
- StarClass: Interactive Visual Classification Using Star Coordinates2003
- Long Short-Term MemoryNeural Computation, 1997