LSTMVis: A Tool for Visual Analysis of Hidden State Dynamics in Recurrent Neural Networks

Abstract

Recurrent neural networks, and in particular long short-term memory (LSTM) networks, are a remarkably effective tool for sequence modeling that learn a dense black-box hidden representation of their sequential input. Researchers interested in better understanding these models have studied the changes in hidden state representations over time and noticed some interpretable patterns but also significant noise. In this work, we present LSTMVIS, a visual analysis tool for recurrent neural networks with a focus on understanding these hidden state dynamics. The tool allows users to select a hypothesis input range to focus on local state changes, to match these states changes to similar patterns in a large data set, and to align these results with structural annotations from their domain. We show several use cases of the tool for analyzing specific hidden state properties on dataset containing nesting, phrase structure, and chord progressions, and demonstrate how the tool can be used to isolate patterns for further statistical analysis. We characterize the domain, the different stakeholders, and their goals and tasks.

InfoVis 2017

References in corpus (9)

• Sequence to Sequence Learning with Neural Networks
• TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
• Recurrent Neural Network Regularization
• Deep Speech 2: End-to-End Speech Recognition in English and Mandarin
• Modeling Temporal Dependencies in High-Dimensional Sequences: Application to Polyphonic Music Generation and Transcription
• Predicting Process Behaviour using Deep Learning
• Modeling Language Vagueness in Privacy Policies using Deep Neural Networks
• rnn : Recurrent Library for Torch
• Representation of linguistic form and function in recurrent neural networks

Cited by in corpus (10)

• Understanding Neural Networks through Representation Erasure
• Beyond Word Importance: Contextual Decomposition to Extract Interactions from LSTMs
• Improving Visualization Interpretation Using Counterfactuals
• Self-Explaining Structures Improve NLP Models
• Interpreting Deep Learning Models in Natural Language Processing: A Review
• Understanding Early Word Learning in Situated Artificial Agents
• Pair the Dots: Jointly Examining Training History and Test Stimuli for Model Interpretability
• X-ToM: Explaining with Theory-of-Mind for Gaining Justified Human Trust
• Automatic Detection of Vague Words and Sentences in Privacy Policies
• XAlgo: a Design Probe of Explaining Algorithms' Internal States via Question-Answering