Modeling Structure‐Building in the Brain With CCG Parsing and Large Language Models
Cognitive Science · 2023-07 · 47 (7)
DOI: 10.1111/cogs.13312 ↗Kurz gefasst
Fachartikel in der Zeitschrift Cognitive Science, online erschienen am 07.07.2023. Verfasst von Miloš Stanojević und 4 weiteren Autor:innen. Beteiligte Einrichtung: Google DeepMind. Hauptthemen laut MEDLINE: Language, Brain. Die Arbeit wurde bisher 43-mal zitiert (Crossref) und nennt 109 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Directorate for Computer and Information Science and Engineering, National Science Foundation.
Abstract (Originalsprache)
To model behavioral and neural correlates of language comprehension in naturalistic environments, researchers have turned to broad‐coverage tools from natural‐language processing and machine learning. Where syntactic structure is explicitly modeled, prior work has relied predominantly on context‐free grammars (CFGs), yet such formalisms are not sufficiently expressive for human languages. Combinatory categorial grammars (CCGs) are sufficiently expressive directly compositional models of grammar with flexible constituency that affords incremental interpretation. In this work, we evaluate whether a more expressive CCG provides a better model than a CFG for human neural signals collected with functional magnetic resonance imaging (fMRI) while participants listen to an audiobook story. We further test between variants of CCG that differ in how they handle optional adjuncts. These evaluations are carried out against a baseline that includes estimates of next‐word predictability from a transformer neural network language model. Such a comparison reveals unique contributions of CCG structure‐building predominantly in the left posterior temporal lobe: CCG‐derived measures offer a superior fit to neural signals compared to those derived from a CFG. These effects are spatially distinct from bilateral superior temporal effects that are unique to predictability. Neural effects for structure‐building are thus separable from predictability during naturalistic listening, and those effects are best characterized by a grammar whose expressive power is motivated on independent linguistic grounds.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Google DeepMind.
- Department of Linguistics, University of Michigan.
- Department of Linguistics, University of Georgia.
- School of Informatics, University of Edinburgh.
ORCID-Profile: Jonathan R Brennan
Förderung
- Directorate for Computer and Information Science and Engineering
- National Science Foundation
Zitieren
APA
Stanojević, M., Brennan, J. R., Dunagan, D., Steedman, M., & Hale, J. T. (2023). Modeling Structure‐Building in the Brain With CCG Parsing and Large Language Models. Cognitive Science, 47(7). https://doi.org/10.1111/cogs.13312
Vancouver
Stanojević M, Brennan JR, Dunagan D, Steedman M, Hale JT. Modeling Structure‐Building in the Brain With CCG Parsing and Large Language Models. Cognitive Science. 2023;47(7). doi:10.1111/cogs.13312
BibTeX
@article{Stanojevi2023,
author = {Stanojević, Miloš and Brennan, Jonathan R and Dunagan, Donald and Steedman, Mark and Hale, John T},
title = {Modeling Structure‐Building in the Brain With CCG Parsing and Large Language Models},
journal = {Cognitive Science},
year = {2023},
volume = {47},
number = {7},
doi = {10.1111/cogs.13312}
}
EUVIMED