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

Optimizing one-shot learning with binary synapses.

Sandro Romani1, Daniel J Amit, Yali Amit

  • 1Human Physiology, Università di Roma La Sapienza, Rome 00185, Italy. sandro.romani@gmail.com

Neural Computation
|April 5, 2008
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

Sexual dimorphism in the complete connectome of the <i>Drosophila</i> male central nervous system.

bioRxiv : the preprint server for biology·2025
Same author

Connectome-driven neural inventory of a complete visual system.

Nature·2025
Same author

Mechanisms of memory-supporting neuronal dynamics in hippocampal area CA3.

Cell·2025
Same author

Mechanisms of memory-supporting neuronal dynamics in hippocampal area CA3.

Cell·2024
Same author

Maintaining and updating accurate internal representations of continuous variables with a handful of neurons.

Nature neuroscience·2024
Same author

Connectome-driven neural inventory of a complete visual system.

bioRxiv : the preprint server for biology·2024
Same journal

A Model-Free Reinforcement Learning Implementation of Decision Making Under Uncertainty by Sequential Sampling.

Neural computation·2026
Same journal

DROP: Distributional and Regular Optimism and Pessimism for Reinforcement Learning.

Neural computation·2026
Same journal

Hierarchical Active Inference Using Successor Representations.

Neural computation·2026
Same journal

W-Kernel and Its Principal Space for Frequentist Evaluation of Bayesian Estimators.

Neural computation·2026
Same journal

A Hidden Markov Model-Inspired Sequence Classification Method for Hyperdimensional Computing.

Neural computation·2026
Same journal

Sparse Graphical Modeling for Electrophysiological Phase-Based Connectivity Using Circular Statistics.

Neural computation·2026
See all related articles

This study models how neural networks recognize familiar stimuli and retrieve memories. It finds that networks can balance familiarity detection and memory recall, with learning speed influencing which function is prioritized.

Area of Science:

  • Computational neuroscience
  • Artificial neural networks
  • Machine learning

Background:

  • Hebbian learning principles are foundational for understanding synaptic plasticity and memory formation.
  • Previous models focused on either memory retrieval or familiarity recognition, but not both simultaneously.
  • Understanding neural network capacity is crucial for developing effective memory models.

Purpose of the Study:

  • To develop a unified computational model for both familiarity recognition and memory retrieval.
  • To analyze the capacity limits of neural networks trained with Hebbian learning.
  • To investigate the impact of learning speed (potentiation probability) on network performance.

Main Methods:

  • Utilized a network of excitatory synapses trained with a conservative Hebbian learning rule.

Related Experiment Videos

  • Extended mathematical analysis (Amit and Fusi, 1994) for binary neurons, focusing on signal-to-noise ratio.
  • Simplified capacity analysis for analog neurons by examining synapse potentiation above steady-state.
  • Corroborated analytical findings with computational simulations.

    • Main Results:

    • Fast learning (high potentiation probability) enables retrieval of recent patterns in working memory.
    • Networks demonstrate high capacity for familiarity recognition of numerous once-seen patterns.
    • Slow learning (low potentiation probability) preserves familiarity recognition but diminishes memory retrieval.
    • Derived optimal constraints for potentiation and depression probabilities in analog neurons.

    Conclusions:

    • A single Hebbian learning framework can effectively model both familiarity recognition and memory retrieval.
    • Learning speed is a critical parameter determining the balance between memory recall and familiarity detection.
    • The model provides insights into the capacity and functional trade-offs of neural memory systems.