Understanding Recurrent Neural State Using Memory Signatures

Abstract

We demonstrate a network visualization technique to analyze the recurrent state inside the LSTMs/GRUs used commonly in language and acoustic models. Interpreting intermediate state and network activations inside end-to-end models remains an open challenge. Our method allows users to understand exactly how much and what history is encoded inside recurrent state in grapheme sequence models. Our procedure trains multiple decoders that predict prior input history. Compiling results from these decoders, a user can obtain a signature of the recurrent kernel that characterizes its memory behavior. We demonstrate this method's usefulness in revealing information divergence in the bases of recurrent factorized kernels, visualizing the character-level differences between the memory of n-gram and recurrent language models, and extracting knowledge of history encoded in the layers of grapheme-based end-to-end ASR networks.

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keykoppula2018understanding

Related papers

• Memory Visualization For Gated Recurrent Neural Networks In Speech Recognition (2016)11.76
• From Nodes To Networks: Evolving Recurrent Neural Networks (2018)0.00
• Frame Stacking And Retaining For Recurrent Neural Network Acoustic Model (2017)0.00
• Analyzing Hidden Representations In End-to-end Automatic Speech Recognition Systems (2017)0.00
• Word Recognition, Competition, And Activation In A Model Of Visually Grounded Speech (2019)0.00
• Persistent Hidden States And Nonlinear Transformation For Long Short-term Memory (2018)8.09
• Incremental Training Of A Recurrent Neural Network Exploiting A Multi-scale Dynamic Memory (2020)3.58
• Investigating Gated Recurrent Neural Networks For Speech Synthesis (2016)0.00