<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="parovic-etal-2023-cross">
    <titleInfo>
        <title>Cross-Lingual Transfer with Target Language-Ready Task Adapters</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Marinela</namePart>
        <namePart type="family">Parovic</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Alan</namePart>
        <namePart type="family">Ansell</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Ivan</namePart>
        <namePart type="family">Vulić</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Anna</namePart>
        <namePart type="family">Korhonen</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2023-07</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Findings of the Association for Computational Linguistics: ACL 2023</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Anna</namePart>
            <namePart type="family">Rogers</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Jordan</namePart>
            <namePart type="family">Boyd-Graber</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Naoaki</namePart>
            <namePart type="family">Okazaki</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Toronto, Canada</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
    </relatedItem>
    <abstract>Adapters have emerged as a modular and parameter-efficient approach to (zero-shot) cross-lingual transfer. The established MAD-X framework employs separate language and task adapters which can be arbitrarily combined to perform the transfer of any task to any target language. Subsequently, BAD-X, an extension of the MAD-X framework, achieves improved transfer at the cost of MAD-X’s modularity by creating ‘bilingual’ adapters specific to the source-target language pair. In this work, we aim to take the best of both worlds by (i) fine-tuning *task* adapters adapted to the target language(s) (so-called *‘target language-ready’ (TLR)* adapters) to maintain high transfer performance, but (ii) without sacrificing the highly modular design of MAD-X. The main idea of ‘target language-ready’ adapters is to resolve the training-vs-inference discrepancy of MAD-X: the task adapter ‘sees’ the target language adapter for the very first time during inference, and thus might not be fully compatible with it. We address this mismatch by exposing the task adapter to the target language adapter during training, and empirically validate several variants of the idea: in the simplest form, we alternate between using the source and target language adapters during task adapter training, which can be generalized to cycling over any set of language adapters. We evaluate different TLR-based transfer configurations with varying degrees of generality across a suite of standard cross-lingual benchmarks, and find that the most general (and thus most modular) configuration consistently outperforms MAD-X and BAD-X on most tasks and languages.</abstract>
    <identifier type="citekey">parovic-etal-2023-cross</identifier>
    <identifier type="doi">10.18653/v1/2023.findings-acl.13</identifier>
    <location>
        <url>https://aclanthology.org/2023.findings-acl.13</url>
    </location>
    <part>
        <date>2023-07</date>
        <extent unit="page">
            <start>176</start>
            <end>193</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Cross-Lingual Transfer with Target Language-Ready Task Adapters
%A Parovic, Marinela
%A Ansell, Alan
%A Vulić, Ivan
%A Korhonen, Anna
%Y Rogers, Anna
%Y Boyd-Graber, Jordan
%Y Okazaki, Naoaki
%S Findings of the Association for Computational Linguistics: ACL 2023
%D 2023
%8 July
%I Association for Computational Linguistics
%C Toronto, Canada
%F parovic-etal-2023-cross
%X Adapters have emerged as a modular and parameter-efficient approach to (zero-shot) cross-lingual transfer. The established MAD-X framework employs separate language and task adapters which can be arbitrarily combined to perform the transfer of any task to any target language. Subsequently, BAD-X, an extension of the MAD-X framework, achieves improved transfer at the cost of MAD-X’s modularity by creating ‘bilingual’ adapters specific to the source-target language pair. In this work, we aim to take the best of both worlds by (i) fine-tuning *task* adapters adapted to the target language(s) (so-called *‘target language-ready’ (TLR)* adapters) to maintain high transfer performance, but (ii) without sacrificing the highly modular design of MAD-X. The main idea of ‘target language-ready’ adapters is to resolve the training-vs-inference discrepancy of MAD-X: the task adapter ‘sees’ the target language adapter for the very first time during inference, and thus might not be fully compatible with it. We address this mismatch by exposing the task adapter to the target language adapter during training, and empirically validate several variants of the idea: in the simplest form, we alternate between using the source and target language adapters during task adapter training, which can be generalized to cycling over any set of language adapters. We evaluate different TLR-based transfer configurations with varying degrees of generality across a suite of standard cross-lingual benchmarks, and find that the most general (and thus most modular) configuration consistently outperforms MAD-X and BAD-X on most tasks and languages.
%R 10.18653/v1/2023.findings-acl.13
%U https://aclanthology.org/2023.findings-acl.13
%U https://doi.org/10.18653/v1/2023.findings-acl.13
%P 176-193

Markdown (Informal)

[Cross-Lingual Transfer with Target Language-Ready Task Adapters](https://aclanthology.org/2023.findings-acl.13) (Parovic et al., Findings 2023)

• Cross-Lingual Transfer with Target Language-Ready Task Adapters (Parovic et al., Findings 2023)

ACL

• Marinela Parovic, Alan Ansell, Ivan Vulić, and Anna Korhonen. 2023. Cross-Lingual Transfer with Target Language-Ready Task Adapters. In Findings of the Association for Computational Linguistics: ACL 2023, pages 176–193, Toronto, Canada. Association for Computational Linguistics.