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

Inferring input nonlinearities in neural encoding models.

Misha B Ahrens1, Liam Paninski, Maneesh Sahani

  • 1Gatsby Computational Neuroscience Unit, University College London, London, UK.

Network (Bristol, England)
|February 27, 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

Beast3D: Animal behavioral analysis and neural encoding from multi-view video via Gaussian splatting.

ArXiv·2026
Same author

POCO: Scalable Neural Forecasting through Population Conditioning.

Advances in neural information processing systems·2026
Same author

Lightning Pose 3D: an uncertainty-aware framework for data-efficient multi-view animal pose estimation.

bioRxiv : the preprint server for biology·2026
Same author

A neuron-glia circuit anticipates hypoxia to regulate organismal oxygen use.

bioRxiv : the preprint server for biology·2026
Same author

Computational optimization of two-photon holographic stimulation sites<i>in vivo</i>.

Journal of neural engineering·2026
Same author

Animal behavioral analysis and neural encoding with transformer-based self-supervised pretraining.

ArXiv·2026
Same journal

Enhancing IoT security: A Creative Swagger Optimization algorithm for DDoS defence.

Network (Bristol, England)·2026
Same journal

Parametric optimization for electrical discharge diamond grinding (EDDG) system using dual approach.

Network (Bristol, England)·2025
Same journal

A novel lung cancer diagnosis model using hybrid convolution (2D/3D)-based adaptive DenseUnet with attention mechanism.

Network (Bristol, England)·2025
Same journal

Hybrid optimization enabled Eff-FDMNet for Parkinson's disease detection and classification in federated learning.

Network (Bristol, England)·2025
Same journal

AI-driven plant disease detection with tailored convolutional neural network.

Network (Bristol, England)·2025
Same journal

Layer modified residual Unet++ for speech enhancement using Aquila Black widow optimizer algorithm.

Network (Bristol, England)·2025
See all related articles

We developed new computational models to predict neuronal firing rates from dynamic stimuli. These models use a learned nonlinear transformation followed by a linear filter, accurately capturing neural responses.

Area of Science:

  • Computational Neuroscience
  • Neural Coding
  • Machine Learning in Neuroscience

Background:

  • Understanding how neurons encode dynamic stimuli is crucial for deciphering brain function.
  • Existing models often struggle to capture the complex nonlinearities in neural processing.
  • Dynamic stimuli require models that can process time-varying input sequences.

Purpose of the Study:

  • To introduce a novel class of models for predicting neuronal firing rates from dynamic stimuli.
  • To develop algorithms for estimating model parameters, including nonlinear transformations and linear filters.
  • To validate the models' predictive power and ability to capture neuronal response properties.

Main Methods:

  • Developed two model architectures: one with a shared input nonlinearity and one with a time-varying nonlinearity.

Related Experiment Videos

  • Implemented algorithms for simultaneous estimation of nonlinear transforms and linear filter weights.
  • Utilized regularization techniques and uncertainty quantification for robust parameter estimation.
  • Validated models on synthetic data and electrophysiological recordings from rodent barrel cortex.

    • Main Results:

    • Both model types accurately predicted neuronal responses to novel stimuli.
    • The models successfully recovered key properties of neuronal responses.
    • The shared input nonlinearity model offers a more parsimonious approach, while the time-varying nonlinearity model provides greater flexibility.
    • Model performance is dependent on the number of degrees of freedom and available data.

    Conclusions:

    • The proposed models provide an effective framework for understanding neural responses to dynamic stimuli.
    • These models can accurately predict neural activity and reveal underlying response mechanisms.
    • The developed methods offer a powerful tool for analyzing neural data in systems neuroscience.