France or Spain or Germany or France: A Neural Account of Non-Redundant Redundant Disjunctions

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Boguraev, Sasha, Yao, Qing, Mahowald, Kyle
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866915982980677632
author Boguraev, Sasha
Yao, Qing
Mahowald, Kyle
author_facet Boguraev, Sasha
Yao, Qing
Mahowald, Kyle
contents Sentences like "She will go to France or Spain, or perhaps to Germany or France." appear formally redundant, yet become acceptable in contexts such as "Mary will go to a philosophy program in France or Spain, or a mathematics program in Germany or France." While this phenomenon has typically been analyzed using symbolic formal representations, we aim to provide an account grounded in artificial neural mechanisms. We first present new behavioral evidence from humans and large language models demonstrating the robustness of this apparent non-redundancy across contexts. We then show that, in language models, redundancy avoidance arises from two interacting mechanisms: models learn to bind contextually relevant information to repeated lexical items, and Transformer induction heads selectively attend to these context-licensed representations. We argue that this neural explanation sheds light on the mechanisms underlying context-sensitive semantic interpretation, and that it complements existing symbolic analyses.
format Preprint
id arxiv_https___arxiv_org_abs_2602_23547
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle France or Spain or Germany or France: A Neural Account of Non-Redundant Redundant Disjunctions
Boguraev, Sasha
Yao, Qing
Mahowald, Kyle
Computation and Language
Sentences like "She will go to France or Spain, or perhaps to Germany or France." appear formally redundant, yet become acceptable in contexts such as "Mary will go to a philosophy program in France or Spain, or a mathematics program in Germany or France." While this phenomenon has typically been analyzed using symbolic formal representations, we aim to provide an account grounded in artificial neural mechanisms. We first present new behavioral evidence from humans and large language models demonstrating the robustness of this apparent non-redundancy across contexts. We then show that, in language models, redundancy avoidance arises from two interacting mechanisms: models learn to bind contextually relevant information to repeated lexical items, and Transformer induction heads selectively attend to these context-licensed representations. We argue that this neural explanation sheds light on the mechanisms underlying context-sensitive semantic interpretation, and that it complements existing symbolic analyses.
title France or Spain or Germany or France: A Neural Account of Non-Redundant Redundant Disjunctions
topic Computation and Language
url https://arxiv.org/abs/2602.23547