arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1606.07461cs.CLcs.AIcs.NE
收藏

LSTMVis:一种循环神经网络隐态动力学的可视化分析工具

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

Harvard School of Engineering and Applied Sciences(哈佛大学工程与应用科学学院)

Hendrik Strobelt, Sebastian Gehrmann, Hanspeter Pfister, Alexander M. Rush

摘要:

循环神经网络,特别是长短期记忆(LSTM)网络,是一种非常有效的序列建模工具,它学习序列输入的密集黑盒隐藏表示。有兴趣更好地理解这些模型的研究人员研究了隐藏状态表示随时间的变化,并注意到一些可解释的模式,但也有显着的噪音。在这项工作中,我们提出了LSTMVIS,一个循环神经网络的可视化分析工具,重点是理解这些隐藏的状态动态。该工具允许用户选择一个假设输入范围,专注于本地状态变化,将这些状态变化与大型数据集中的类似模式相匹配,并将这些结果与其域的结构注释对齐。我们展示了该工具的几个用例,用于分析包含嵌套,短语结构和和弦进行的数据集上的特定隐藏状态属性,并演示了该工具如何用于隔离模式以进行进一步的统计分析。我们描述了领域,不同的利益相关者,以及他们的目标和任务。

英文摘要:

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.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑