AutoencoderComplexity

This repository contains a simple autoencoder on a Wikipedia sentences dataset to test the idea of whether an autoencoder can be indicative of the complexity of text by exploiting compression complexity.