Neuromodulators Generate Multiple Context-Relevant Behaviors in Recurrent Neural Networks
Neural Computation · 27.02.2026 · 38 (3) · 292-327
DOI: 10.1162/neco.a.1489 ↗Kurz gefasst
Fachartikel in der Zeitschrift Neural Computation, online erschienen am 27.02.2026. Verfasst von Ben Tsuda und 4 weiteren Autor:innen. Beteiligte Einrichtung: Computational Neurobiology Laboratory, Salk Institute for Biological Studies, La Jolla, CA 92037, USA btsuda@mgh.harvard.edu. Hauptthemen laut MEDLINE: Neurons, Neurotransmitter Agents, Models, Neurological, Neural Networks, Computer. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 84 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: NCCIH NIH HHS, NINDS NIH HHS, NIMH NIH HHS.
Abstract (Originalsprache)
Neuromodulators are critical controllers of neural states, with dysfunctions linked to various neuropsychiatric disorders. Although many biological aspects of neuromodulation have been studied, the computational principles underlying how neuromodulation of distributed neural populations controls brain states remain unclear. In contrast to external contextual inputs, neuromodulation can act as a single scalar signal that is broadcast to a vast population of neurons. We model the modulation of synaptic weight in a recurrent neural network model and show that neuromodulators can dramatically alter the function of a network, even when highly simplified. We find that under structural constraints like those in brains, this provides a fundamental mechanism that can increase the computational capability and flexibility of a neural network. Diffuse synaptic weight modulation enables storage of multiple memories using a common set of synapses that are able to generate diverse, even diametrically opposed, behaviors. Our findings help explain how neuromodulators unlock specific behaviors by creating task-specific hyperchannels in neural activity space and motivate more flexible, compact and capable machine learning architectures.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Substanzen
Autor:innen und Institutionen
- Computational Neurobiology Laboratory, Salk Institute for Biological Studies, La Jolla, CA 92037, USA btsuda@mgh.harvard.edu.
- Computational Neurobiology Laboratory, Salk Institute for Biological Studies, La Jolla, CA 92037, USA stefan.pate@northwestern.edu.
- Systems Neuroscience Laboratory, Salk Institute for Biological Studies, La Jolla, CA 92037, USA tye@salk.edu.
- Biologically Inspired Neural and Dynamical Systems Laboratory, School of Computer Science, University of Massachusetts Amherst, Amherst, MA, 01003 hava@umass.edu.
- Computational Neurobiology Laboratory, Salk Institute for Biological Studies, La Jolla, CA 92037, USA.
- Institute for Neural Computation and Division of Biological Sciences, University of California San Diego, La Jolla, CA 92093, USA terry@salk.edu.
Förderung
- NCCIH NIH HHS
- NINDS NIH HHS
- NIMH NIH HHS
Zitieren
APA
Tsuda, B., Pate, S. C., Tye, K. M., Siegelmann, H. T., & Sejnowski, T. J. (2026). Neuromodulators Generate Multiple Context-Relevant Behaviors in Recurrent Neural Networks. Neural Computation, 38(3), 292-327. https://doi.org/10.1162/neco.a.1489
Vancouver
Tsuda B, Pate SC, Tye KM, Siegelmann HT, Sejnowski TJ. Neuromodulators Generate Multiple Context-Relevant Behaviors in Recurrent Neural Networks. Neural Computation. 2026;38(3):292-327. doi:10.1162/neco.a.1489
BibTeX
@article{Tsuda2026,
author = {Tsuda, Ben and Pate, Stefan C and Tye, Kay M and Siegelmann, Hava T and Sejnowski, Terrence J},
title = {Neuromodulators Generate Multiple Context-Relevant Behaviors in Recurrent Neural Networks},
journal = {Neural Computation},
year = {2026},
volume = {38},
number = {3},
pages = {292--327},
doi = {10.1162/neco.a.1489}
}
EUVIMED