@inproceedings{araujo-etal-2021-augmenting,
title = "Augmenting {BERT}-style Models with Predictive Coding to Improve Discourse-level Representations",
author = "Araujo, Vladimir and
Villa, Andr{\'e}s and
Mendoza, Marcelo and
Moens, Marie-Francine and
Soto, Alvaro",
booktitle = "Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing",
month = nov,
year = "2021",
address = "Online and Punta Cana, Dominican Republic",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2021.emnlp-main.240",
doi = "10.18653/v1/2021.emnlp-main.240",
pages = "3016--3022",
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.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="araujo-etal-2021-augmenting">
<titleInfo>
<title>Augmenting BERT-style Models with Predictive Coding to Improve Discourse-level Representations</title>
</titleInfo>
<name type="personal">
<namePart type="given">Vladimir</namePart>
<namePart type="family">Araujo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Andrés</namePart>
<namePart type="family">Villa</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Marcelo</namePart>
<namePart type="family">Mendoza</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Marie-Francine</namePart>
<namePart type="family">Moens</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alvaro</namePart>
<namePart type="family">Soto</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2021-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Online and Punta Cana, Dominican Republic</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<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.</abstract>
<identifier type="citekey">araujo-etal-2021-augmenting</identifier>
<identifier type="doi">10.18653/v1/2021.emnlp-main.240</identifier>
<location>
<url>https://aclanthology.org/2021.emnlp-main.240</url>
</location>
<part>
<date>2021-11</date>
<extent unit="page">
<start>3016</start>
<end>3022</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Augmenting BERT-style Models with Predictive Coding to Improve Discourse-level Representations
%A Araujo, Vladimir
%A Villa, Andrés
%A Mendoza, Marcelo
%A Moens, Marie-Francine
%A Soto, Alvaro
%S Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing
%D 2021
%8 November
%I Association for Computational Linguistics
%C Online and Punta Cana, Dominican Republic
%F araujo-etal-2021-augmenting
%X 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.
%R 10.18653/v1/2021.emnlp-main.240
%U https://aclanthology.org/2021.emnlp-main.240
%U https://doi.org/10.18653/v1/2021.emnlp-main.240
%P 3016-3022
Markdown (Informal)
[Augmenting BERT-style Models with Predictive Coding to Improve Discourse-level Representations](https://aclanthology.org/2021.emnlp-main.240) (Araujo et al., EMNLP 2021)
ACL