Infusing Finetuning with Semantic Dependencies

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F21%3A10441618" target="_blank" >RIV/00216208:11320/21:10441618 - isvavai.cz</a>

Result on the web

<a href="https://verso.is.cuni.cz/pub/verso.fpl?fname=obd_publikace_handle&handle=.a8a-gSDHQ" target="_blank" >https://verso.is.cuni.cz/pub/verso.fpl?fname=obd_publikace_handle&handle=.a8a-gSDHQ</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1162/tacl_a_00363" target="_blank" >10.1162/tacl_a_00363</a>

Result language

angličtina

Original language name

Infusing Finetuning with Semantic Dependencies

Original language description

For natural language processing systems, two kinds of evidence support the use of text representations from neural language models &quot;pretrained&apos;&apos; on large unannotated corpora: performance on application-inspired benchmarks (Peters et al., 2018, inter alia), and the emergence of syntactic abstractions in those representations (Tenney et al., 2019, inter alia). On the other hand, the lack of grounded supervision calls into question how well these representations can ever capture meaning (Bender and Koller, 2020). We apply novel probes to recent language models-specifically focusing on predicate-argument structure as operationalized by semantic dependencies (Ivanova et al., 2012)-and find that, unlike syntax, semantics is not brought to the surface by today&apos;s pretrained models. We then use convolutional graph encoders to explicitly incorporate semantic parses into task-specific finetuning, yielding benefits to natural language understanding (NLU) tasks in the GLUE benchmark. This approach demonstrates the potential for general-purpose (rather than task-specific) linguistic supervision, above and beyond conventional pretraining and finetuning. Several diagnostics help to localize the benefits of our approach.(1)

Czech name

—

Czech description

—

Type

J_imp - Article in a specialist periodical, which is included in the Web of Science database

CEP classification

—

OECD FORD branch

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Project

—

Continuities

—

Publication year

2021

Confidentiality

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Name of the periodical

Transactions of the Association for Computational Linguistics [online]

ISSN

2307-387X

e-ISSN

—

Volume of the periodical

9

Issue of the periodical within the volume

11.03.2021

Country of publishing house

US - UNITED STATES

Number of pages

17

Pages from-to

226-242

UT code for WoS article

000751952200014

EID of the result in the Scopus database

2-s2.0-85107894350