Skip to content
DraconiticResearch
Search Sources API
201,683 studies31 sourcesUpdated 22 min ago
Peer-reviewedCC BYNature

A critical initialization for biological neural networks

  • Marius Pachitariu1,2
  • Lin Zhong3,2
  • Alexa L. Gracias3,2
  • Amanda Minisi3,2
  • Crystall Lopez3,2
  • Carsen Stringer4,2
4 affiliations
  1. 1HHMI Janelia Research Campus, Ashburn, VA, USA. [email protected]
  2. 2HHMI Janelia Research Campus
  3. 3HHMI Janelia Research Campus, Ashburn, VA, USA
  4. 4HHMI Janelia Research Campus, Ashburn, VA, USA. [email protected]
Open at sourceDOICite

Abstract

Intrinsically generated, brainwide neural activity displays macroscopic coordination among large populations of neurons that persists beyond the biophysical timescales of individual neurons 1–3 . It is not well understood how these macroscopic behaviours arise from microscopic, short-lived interactions between pairs of neurons. Here we show that the eigenvalue spectrum and dynamical properties of large-scale neural recordings in mice are similar to those produced by linear dynamics governed by a random symmetric matrix that is critically normalized. An exception was population activity in hippocampal area CA1, which resembled an efficient, uncorrelated neural code that may be optimized for information storage capacity. High-dimensional, global activity modes emerged in critically normalized artificial networks and persisted under sparse, clustered or spatial connectivity. These dynamics were useful for solving time-dependent tasks such as a zero-shot working memory task.

Related

  1. Harnessing the Intrinsic Dynamics of Biological Neural Networks for Reservoir Computing
    S. Somathilaka, J. Clouse, S. Balasubramaniam
    bioRxivPreprintCC BY-NCFull textNeurosciencebioRxiv
  2. "More Is Different'' in Neural Circuits: Algebraic Emergence of Effective Theories in Canonical Recurrent Motifs of Biological Neuronal Networks
    Nima Dehghani
    arXivPreprintCC BY 4.0Full textNeuroscienceBiological & medical physicsarXiv
  3. Heterogeneous graph neural networks with biological prior knowledge for interpretable drug repurposing in triple-negative breast cancer
    C. Fernandez-Lozano, D. Ferreiro, P. V.-del-Rio
    bioRxivPreprintCC BYFull textBioinformatics & computational biologybioRxiv
  4. Plant–Insect Interactions in Biological Invasions: From Classical Hypotheses to Multitrophic Networks and Biological Control
    Qinchun Li, Fanping Xu, Wenzhi Cheng +2
    InsectsPeer-reviewedCC BYFull textGeneral biologyEurope PMC
  5. Cellular diversity in neural organoids is critical for neurodegenerative disease modeling
    Shiying Liu, Shi-Yan Ng
    PLOS BiologyPeer-reviewedCC BY 4.0Full textCell biologyDevelopmental biologyPLOSPMC

Cite

BibTeXRIS
BibTeX
@article{pachitariu2026critical,
  title = {{A critical initialization for biological neural networks}},
  author = {Marius Pachitariu and Lin Zhong and Alexa L. Gracias and Amanda Minisi and Crystall Lopez and Carsen Stringer},
  journal = {Nature},
  year = {2026},
  month = may,
  doi = {10.1038/s41586-026-10528-1},
  url = {https://doi.org/10.1038/s41586-026-10528-1},
  note = {CC BY},
}
RIS
TY  - JOUR
TI  - A critical initialization for biological neural networks
AU  - Marius Pachitariu
AU  - Lin Zhong
AU  - Alexa L. Gracias
AU  - Amanda Minisi
AU  - Crystall Lopez
AU  - Carsen Stringer
PY  - 2026
DA  - 2026/05/20
JO  - Nature
DO  - 10.1038/s41586-026-10528-1
UR  - https://doi.org/10.1038/s41586-026-10528-1
AB  - Intrinsically generated, brainwide neural activity displays macroscopic coordination among large populations of neurons that persists beyond the biophysical timescales of individual neurons 1–3 . It is not well understood how these macroscopic behaviours arise from microscopic, short-lived interactions between pairs of neurons. Here we show that the eigenvalue spectrum and dynamical properties of large-scale neural recordings in mice are similar to those produced by linear dynamics governed by a random symmetric matrix that is critically normalized. An exception was population activity in hippocampal area CA1, which resembled an efficient, uncorrelated neural code that may be optimized for information storage capacity. High-dimensional, global activity modes emerged in critically normalized artificial networks and persisted under sparse, clustered or spatial connectivity. These dynamics were useful for solving time-dependent tasks such as a zero-shot working memory task.
SN  - 0028-0836
SN  - 1476-4687
ER  -