# Explaining Recurrent Neural Network Predictions in Sentiment Analysis

## TL;DR

This paper extends Layer-wise Relevance Propagation (LRP) to recurrent neural networks like LSTMs and GRUs, providing more insightful explanations for sentiment analysis predictions and outperforming gradient-based methods.

## Contribution

We develop a specific LRP propagation rule for recurrent architectures and demonstrate its effectiveness on sentiment analysis tasks.

## Key findings

- LRP provides more insightful explanations than gradient-based methods.
- The proposed method improves relevance attribution quality.
- Relevance scores align well with human intuition.

## Abstract

Recently, a technique called Layer-wise Relevance Propagation (LRP) was shown to deliver insightful explanations in the form of input space relevances for understanding feed-forward neural network classification decisions. In the present work, we extend the usage of LRP to recurrent neural networks. We propose a specific propagation rule applicable to multiplicative connections as they arise in recurrent network architectures such as LSTMs and GRUs. We apply our technique to a word-based bi-directional LSTM model on a five-class sentiment prediction task, and evaluate the resulting LRP relevances both qualitatively and quantitatively, obtaining better results than a gradient-based related method which was used in previous work.