@inproceedings{wilcox-etal-2021-targeted,
title = "A Targeted Assessment of Incremental Processing in Neural Language Models and Humans",
author = "Wilcox, Ethan and
Vani, Pranali and
Levy, Roger",
editor = "Zong, Chengqing and
Xia, Fei and
Li, Wenjie and
Navigli, Roberto",
booktitle = "Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)",
month = aug,
year = "2021",
address = "Online",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2021.acl-long.76",
doi = "10.18653/v1/2021.acl-long.76",
pages = "939--952",
abstract = "We present a targeted, scaled-up comparison of incremental processing in humans and neural language models by collecting by-word reaction time data for sixteen different syntactic test suites across a range of structural phenomena. Human reaction time data comes from a novel online experimental paradigm called the Interpolated Maze task. We compare human reaction times to by-word probabilities for four contemporary language models, with different architectures and trained on a range of data set sizes. We find that across many phenomena, both humans and language models show increased processing difficulty in ungrammatical sentence regions with human and model {`}accuracy{'} scores a la Marvin and Linzen (2018) about equal. However, although language model outputs match humans in direction, we show that models systematically under-predict the difference in magnitude of incremental processing difficulty between grammatical and ungrammatical sentences. Specifically, when models encounter syntactic violations they fail to accurately predict the longer reading times observed in the human data. These results call into question whether contemporary language models are approaching human-like performance for sensitivity to syntactic violations.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wilcox-etal-2021-targeted">
<titleInfo>
<title>A Targeted Assessment of Incremental Processing in Neural Language Models and Humans</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ethan</namePart>
<namePart type="family">Wilcox</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Pranali</namePart>
<namePart type="family">Vani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Roger</namePart>
<namePart type="family">Levy</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2021-08</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Chengqing</namePart>
<namePart type="family">Zong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Fei</namePart>
<namePart type="family">Xia</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wenjie</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Roberto</namePart>
<namePart type="family">Navigli</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Online</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>We present a targeted, scaled-up comparison of incremental processing in humans and neural language models by collecting by-word reaction time data for sixteen different syntactic test suites across a range of structural phenomena. Human reaction time data comes from a novel online experimental paradigm called the Interpolated Maze task. We compare human reaction times to by-word probabilities for four contemporary language models, with different architectures and trained on a range of data set sizes. We find that across many phenomena, both humans and language models show increased processing difficulty in ungrammatical sentence regions with human and model ‘accuracy’ scores a la Marvin and Linzen (2018) about equal. However, although language model outputs match humans in direction, we show that models systematically under-predict the difference in magnitude of incremental processing difficulty between grammatical and ungrammatical sentences. Specifically, when models encounter syntactic violations they fail to accurately predict the longer reading times observed in the human data. These results call into question whether contemporary language models are approaching human-like performance for sensitivity to syntactic violations.</abstract>
<identifier type="citekey">wilcox-etal-2021-targeted</identifier>
<identifier type="doi">10.18653/v1/2021.acl-long.76</identifier>
<location>
<url>https://aclanthology.org/2021.acl-long.76</url>
</location>
<part>
<date>2021-08</date>
<extent unit="page">
<start>939</start>
<end>952</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T A Targeted Assessment of Incremental Processing in Neural Language Models and Humans
%A Wilcox, Ethan
%A Vani, Pranali
%A Levy, Roger
%Y Zong, Chengqing
%Y Xia, Fei
%Y Li, Wenjie
%Y Navigli, Roberto
%S Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)
%D 2021
%8 August
%I Association for Computational Linguistics
%C Online
%F wilcox-etal-2021-targeted
%X We present a targeted, scaled-up comparison of incremental processing in humans and neural language models by collecting by-word reaction time data for sixteen different syntactic test suites across a range of structural phenomena. Human reaction time data comes from a novel online experimental paradigm called the Interpolated Maze task. We compare human reaction times to by-word probabilities for four contemporary language models, with different architectures and trained on a range of data set sizes. We find that across many phenomena, both humans and language models show increased processing difficulty in ungrammatical sentence regions with human and model ‘accuracy’ scores a la Marvin and Linzen (2018) about equal. However, although language model outputs match humans in direction, we show that models systematically under-predict the difference in magnitude of incremental processing difficulty between grammatical and ungrammatical sentences. Specifically, when models encounter syntactic violations they fail to accurately predict the longer reading times observed in the human data. These results call into question whether contemporary language models are approaching human-like performance for sensitivity to syntactic violations.
%R 10.18653/v1/2021.acl-long.76
%U https://aclanthology.org/2021.acl-long.76
%U https://doi.org/10.18653/v1/2021.acl-long.76
%P 939-952
Markdown (Informal)
[A Targeted Assessment of Incremental Processing in Neural Language Models and Humans](https://aclanthology.org/2021.acl-long.76) (Wilcox et al., ACL-IJCNLP 2021)
ACL