Augmenting BERT-style Models with Predictive Coding to Improve Discourse-level Representations
Abstract
Current language models are usually trained using a self-supervised scheme, where the main focus is learning representations at the word or sentence level. However, there has been limited progress in generating useful discourse-level representations. In this work, we propose to use ideas from predictive coding theory to augment BERT-style language models with a mechanism that allows them to learn suitable discourse-level representations. As a result, our proposed approach is able to predict future sentences using explicit top-down connections that operate at the intermediate layers of the network. By experimenting with benchmarks designed to evaluate discourse-related knowledge using pre-trained sentence representations, we demonstrate that our approach improves performance in 6 out of 11 tasks by excelling in discourse relationship detection.
Cite
@article{arxiv.2109.04602,
title = {Augmenting BERT-style Models with Predictive Coding to Improve Discourse-level Representations},
author = {Vladimir Araujo and Andrés Villa and Marcelo Mendoza and Marie-Francine Moens and Alvaro Soto},
journal= {arXiv preprint arXiv:2109.04602},
year = {2021}
}
Comments
Accepted paper EMNLP2021