Academic Journal
Unifying the structures of language in a neural population code.
| Τίτλος: | Unifying the structures of language in a neural population code. |
|---|---|
| Συγγραφείς: | Nastase SA; Department of Psychology and Center for Computational Language Sciences, University of Southern California, Los Angeles, CA, USA. Electronic address: snastase@usc.edu., Zada Z; Princeton Neuroscience Institute, Princeton University, Princeton, NJ, USA; Department of Psychology, Princeton University, Princeton, NJ, USA., Goldberg A; Department of Psychology, Princeton University, Princeton, NJ, USA., Hasson U; Princeton Neuroscience Institute, Princeton University, Princeton, NJ, USA; Department of Psychology, Princeton University, Princeton, NJ, USA. |
| Πηγή: | Neuron [Neuron] 2026 Oct 07; Vol. 114 (19), pp. 3527-3549. Date of Electronic Publication: 2026 Aug 14. |
| Τύπος έκδοσης: | Journal Article; Review |
| Γλώσσα: | English |
| Στοιχεία περιοδικού: | Publisher: Cell Press Country of Publication: United States NLM ID: 8809320 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 1097-4199 (Electronic) Linking ISSN: 08966273 NLM ISO Abbreviation: Neuron Subsets: MEDLINE |
| Imprint Name(s): | Original Publication: [Cambridge, Mass. : Cell Press, c1988- |
| Ιατρικοί όροι (MeSH): | Brain*/physiology , Language* , Models, Neurological*, Humans ; Large Language Models |
| Περίληψη: | Large language models (LLMs) have mastered human language in ways that no previous computational system has. While rule-based, symbolic systems sufficed for constrained, well-defined problems, they were not able to accommodate the context-sensitive expressivity of natural language. LLMs instead use statistical learning to encode the diversity of linguistic structures into a unified high-dimensional embedding space. Strikingly, this context-driven, distributed representation closely parallels neural population codes, suggesting that the human language system may have converged on a similar computational strategy. Drawing on a growing body of work at the intersection of artificial intelligence and cognitive neuroscience, we show that LLMs can serve as cognitively plausible models of the neural computations supporting language in the human brain. We conclude that explaining how language can emerge from neural population codes, in both biological and artificial systems, will not be achieved through the incremental refinement of algebraic-symbolic theories but will demand new theoretical paradigms. (Copyright © 2026 The Authors. Published by Elsevier Inc. All rights reserved.) |
| Competing Interests: | Declaration of interests The authors declare no competing interests. |
| Grant Information: | R01 DC022534 United States DC NIDCD NIH HHS |
| Contributed Indexing: | Keywords: LLMs; connectionism; construction grammar; embedding space; large language models; neurolinguistics; representational geometry; statistical learning |
| Entry Date(s): | Date Created: 20260814 Date Completed: 20261007 Latest Revision: 20261007 |
| Update Code: | 20261008 |
| PubMed Central ID: | PMC13546754 |
| DOI: | 10.1016/j.neuron.2026.07.024 |
| PMID: | 42600614 |
| Βάση Δεδομένων: | MEDLINE |
| ISSN: | 1097-4199 |
|---|---|
| DOI: | 10.1016/j.neuron.2026.07.024 |