Focus shifts in contextual and lexical cue interactions in GPT models
PLOS One · 18.06.2026 · 21 (6) · e0351924
DOI: 10.1371/journal.pone.0351924 ↗Kurz gefasst
Fachartikel in der Zeitschrift PLOS One, online erschienen am 18.06.2026. Verfasst von Wonil Chung und 1 weiteren Autor:innen. Beteiligte Einrichtung: Department of English Language and Literature, Dongguk University, Seoul, South Korea. Hauptthemen laut MEDLINE: Generative Artificial Intelligence, Linguistics. Die Arbeit nennt 75 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Ministry of Education of the Republic of Korea and the National Research Foundation of Korea.
Abstract (Originalsprache)
Transformer-based language models have demonstrated sensitivity to a range of linguistic dependencies, yet it remains unclear how they represent information-structural focus and integrate discourse and lexical focus cues during ellipsis resolution. We investigated GPT-style models’ interpretation of elliptical remnant continuations in double-object constructions by manipulating contextual focus via preceding interrogatives ( who vs. what ) and lexical focus via the particle only , whose surface position was varied. Using word-by-word surprisal as an index of processing difficulty, we conducted three experiments with GPT-2 models (Small–XL) and GPT-Neo. In Experiment 1 (no only ), models robustly tracked the wh -induced discourse focus, assigning higher surprisal to remnants that mismatched the contextually focused constituent. In Experiment 2 ( only preceding the indirect object), contextual focus continued to dominate, indicating that discourse cues were maintained despite the presence of a competing lexical marker. In Experiment 3 ( only preceding the direct object), lexical focus effects became stronger: models favored remnants aligned with the lexically biased direct object, consistent with locality-based cue weighting when only is adjacent to that object. Comparisons with human reaction-time data revealed broad convergence in contextual-focus sensitivity but divergence when the remnant was compatible with one cue but not the other, with GPT-style models exhibiting a stronger bias toward alignment with only than humans. Together, these findings suggest that the tested models maintain discourse-level focus representations while integrating multiple focus cues in a proximity-sensitive manner, revealing both overlap and limits in their alignment with human processing.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Department of English Language and Literature, Dongguk University, Seoul, South Korea.
ORCID-Profile: Keonwoo Koo
Förderung
- Ministry of Education of the Republic of Korea and the National Research Foundation of Korea
Zitieren
APA
Chung, W., & Koo, K. (2026). Focus shifts in contextual and lexical cue interactions in GPT models. PLOS One, 21(6), e0351924. https://doi.org/10.1371/journal.pone.0351924
Vancouver
Chung W, Koo K. Focus shifts in contextual and lexical cue interactions in GPT models. PLOS One. 2026;21(6):e0351924. doi:10.1371/journal.pone.0351924
BibTeX
@article{Chung2026,
author = {Chung, Wonil and Koo, Keonwoo},
title = {Focus shifts in contextual and lexical cue interactions in GPT models},
journal = {PLOS One},
year = {2026},
volume = {21},
number = {6},
pages = {e0351924},
doi = {10.1371/journal.pone.0351924}
}
EUVIMED