<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wu-etal-2023-transparency">
    <titleInfo>
        <title>Transparency Helps Reveal When Language Models Learn Meaning</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Zhaofeng</namePart>
        <namePart type="family">Wu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">William</namePart>
        <namePart type="family">Merrill</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Hao</namePart>
        <namePart type="family">Peng</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Iz</namePart>
        <namePart type="family">Beltagy</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Noah</namePart>
        <namePart type="given">A</namePart>
        <namePart type="family">Smith</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2023</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <genre authority="bibutilsgt">journal article</genre>
    <relatedItem type="host">
        <titleInfo>
            <title>Transactions of the Association for Computational Linguistics</title>
        </titleInfo>
        <originInfo>
            <issuance>continuing</issuance>
            <publisher>MIT Press</publisher>
            <place>
                <placeTerm type="text">Cambridge, MA</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">periodical</genre>
        <genre authority="bibutilsgt">academic journal</genre>
    </relatedItem>
    <abstract>Many current NLP systems are built from language models trained to optimize unsupervised objectives on large amounts of raw text. Under what conditions might such a procedure acquire meaning? Our systematic experiments with synthetic data reveal that, with languages where all expressions have context-independent denotations (i.e., languages with strong transparency), both autoregressive and masked language models successfully learn to emulate semantic relations between expressions. However, when denotations are changed to be context-dependent with the language otherwise unmodified, this ability degrades. Turning to natural language, our experiments with a specific phenomenon—referential opacity—add to the growing body of evidence that current language models do not represent natural language semantics well. We show this failure relates to the context-dependent nature of natural language form-meaning mappings.</abstract>
    <identifier type="citekey">wu-etal-2023-transparency</identifier>
    <identifier type="doi">10.1162/tacl_a_00565</identifier>
    <location>
        <url>https://aclanthology.org/2023.tacl-1.36</url>
    </location>
    <part>
        <date>2023</date>
        <detail type="volume"><number>11</number></detail>
        <extent unit="page">
            <start>617</start>
            <end>634</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Journal Article
%T Transparency Helps Reveal When Language Models Learn Meaning
%A Wu, Zhaofeng
%A Merrill, William
%A Peng, Hao
%A Beltagy, Iz
%A Smith, Noah A.
%J Transactions of the Association for Computational Linguistics
%D 2023
%V 11
%I MIT Press
%C Cambridge, MA
%F wu-etal-2023-transparency
%X Many current NLP systems are built from language models trained to optimize unsupervised objectives on large amounts of raw text. Under what conditions might such a procedure acquire meaning? Our systematic experiments with synthetic data reveal that, with languages where all expressions have context-independent denotations (i.e., languages with strong transparency), both autoregressive and masked language models successfully learn to emulate semantic relations between expressions. However, when denotations are changed to be context-dependent with the language otherwise unmodified, this ability degrades. Turning to natural language, our experiments with a specific phenomenon—referential opacity—add to the growing body of evidence that current language models do not represent natural language semantics well. We show this failure relates to the context-dependent nature of natural language form-meaning mappings.
%R 10.1162/tacl_a_00565
%U https://aclanthology.org/2023.tacl-1.36
%U https://doi.org/10.1162/tacl_a_00565
%P 617-634

Markdown (Informal)

[Transparency Helps Reveal When Language Models Learn Meaning](https://aclanthology.org/2023.tacl-1.36) (Wu et al., TACL 2023)

• Transparency Helps Reveal When Language Models Learn Meaning (Wu et al., TACL 2023)

ACL

• Zhaofeng Wu, William Merrill, Hao Peng, Iz Beltagy, and Noah A. Smith. 2023. Transparency Helps Reveal When Language Models Learn Meaning. Transactions of the Association for Computational Linguistics, 11:617–634.