Response Characterization For Auditing Cell Dynamics In Long Shortterm Memory Networks

# Response Characterization for Auditing Cell Dynamics in Long Short-term Memory Networks

**arXiv ID:** 1809.03864
**Authors:** Ramin M. Hasani, Alexander Amini, Mathias Lechner, Felix Naser, Radu Grosu, Daniela Rus
**Published:** 2018-09-11T13:27:36Z
**Abstract:**
In this paper, we introduce a novel method to interpret recurrent neural networks (RNNs), particularly long short-term memory networks (LSTMs) at the cellular level. We propose a systematic pipeline for interpreting individual hidden state dynamics within the network using response characterization methods. The ranked contribution of individual cells to the network's output is computed by analyzing a set of interpretable metrics of their decoupled step and sinusoidal responses. As a result, our method is able to uniquely identify neurons with insightful dynamics, quantify relationships between dynamical properties and test accuracy through ablation analysis, and interpret the impact of network capacity on a network's dynamical distribution. Finally, we demonstrate generalizability and scalability of our method by evaluating a series of different benchmark sequential datasets.

## Skill Description

This skill is generated from the arXiv paper: Response Characterization for Auditing Cell Dynamics in Long Short-term Memory Networks (1809.03864).

## How to Use

[To be filled in by the user or by future automation]

## References

- [arXiv:1809.03864](http://arxiv.org/abs/1809.03864v1)