Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Optimal computation with attractor networks.

Peter E Latham1, Sophie Deneve, Alexandre Pouget

  • 1Department of Neurobiology, University of California at Los Angeles, Los Angeles, CA 90095-1763, USA. pel@gatsby.ucl.ac.uk

Journal of Physiology, Paris
|July 10, 2004
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Exploiting correlations across trials and behavioral sessions to improve neural decoding.

Neuron·2025
Same author

Brain-wide representations of prior information in mouse decision-making.

Nature·2025
Same author

A brain-wide map of neural activity during complex behaviour.

Nature·2025
Same author

Multi-timescale reinforcement learning in the brain.

Nature·2025
Same author

Reticulon-dependent ER-phagy mediates adaptation to heat stress in C. elegans.

Current biology : CB·2025
Same author

Acquiring musculoskeletal skills with curriculum-based reinforcement learning.

Neuron·2024
Same journal

Role of synchronized physiological and interpersonal rhythms in typical and atypical development.

Journal of physiology, Paris·2017
Same journal

Suicide attempts in children and adolescents: The place of clock genes and early rhythm dysfunction.

Journal of physiology, Paris·2017
Same journal

Editorial.

Journal of physiology, Paris·2017
Same journal

Dyssynchrony and perinatal psychopathology impact of child disease on parents-child interactions, the paradigm of Prader Willi syndrom.

Journal of physiology, Paris·2017
Same journal

Key considerations in designing a speech brain-computer interface.

Journal of physiology, Paris·2017
Same journal

Links between early child maltreatment, mental disorders, and cortisol secretion anomalies.

Journal of physiology, Paris·2017
See all related articles

Multi-dimensional attractor networks can perform reliable computations with noisy neural population codes. Optimal performance is achievable if noise is sufficiently small, depending on its correlational structure.

Area of Science:

  • Computational neuroscience
  • Neural networks
  • Information theory

Background:

  • Neural population codes are essential for brain function.
  • These codes are inherently noisy, posing challenges for reliable computation.
  • Attractor networks are a theoretical model for neural computation.

Purpose of the Study:

  • To investigate the computational reliability of multi-dimensional attractor networks with noisy population codes.
  • To determine the conditions under which these networks can achieve optimal performance.
  • To assess the biological plausibility of such optimal computation in the cortex.

Main Methods:

  • Theoretical analysis of multi-dimensional attractor networks.
  • Investigation of the Cramér-Rao bound for noisy population codes.

Related Experiment Videos

  • Analysis of the impact of noise correlational structure on computational reliability.

    • Main Results:

    • Attractor networks can achieve the Cramér-Rao bound for reliable computation when noise is sufficiently small.
    • The threshold for "small enough" noise depends critically on its correlational structure.
    • Biologically plausible noise levels in the cortex often allow for optimal computation in these networks.

    Conclusions:

    • Multi-dimensional attractor networks offer a framework for understanding reliable neural computation despite noise.
    • Noise correlational structure is a key factor determining the limits of neural computation.
    • These findings have implications for understanding sensory processing and decision-making in biological systems.