Dataset Information

Columnar architecture improves noise robustness in a model cortical network.

ABSTRACT: Cortical columnar architecture was discovered decades ago yet there is no agreed upon explanation for its function. Indeed, some have suggested that it has no function, it is simply an epiphenomenon of developmental processes. To investigate this problem we have constructed a computer model of one square millimeter of layer 2/3 of the primary visual cortex (V1) of the cat. Model cells are connected according to data from recent paired cell studies, in particular the connection probability between pyramidal cells is inversely proportional both to the distance separating the cells and to the distance between the preferred parameters (features) of the cells. We find that these constraints, together with a columnar architecture, produce more tightly clustered populations of cells when compared to the random architecture seen in, for example, rodents. This causes the columnar network to converge more quickly and accurately on the pattern representing a particular stimulus in the presence of noise, suggesting that columnar connectivity functions to improve pattern recognition in cortical circuits. The model also suggests that synaptic failure, a phenomenon exhibited by weak synapses, may conserve metabolic resources by reducing transmitter release at these connections that do not contribute to network function.

SUBMITTER: Bush PC

PROVIDER: S-EPMC4364615 | biostudies-literature | 2015

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Columnar architecture improves noise robustness in a model cortical network.

Bush Paul C PC Mainen Zachary F ZF

PloS one 20150317 3

Cortical columnar architecture was discovered decades ago yet there is no agreed upon explanation for its function. Indeed, some have suggested that it has no function, it is simply an epiphenomenon of developmental processes. To investigate this problem we have constructed a computer model of one square millimeter of layer 2/3 of the primary visual cortex (V1) of the cat. Model cells are connected according to data from recent paired cell studies, in particular the connection probability betwee ...[more]

PMID: 25781314

Similar Datasets

Project description:Task-optimized convolutional neural networks (CNNs) show striking similarities to the ventral visual stream. However, human-imperceptible image perturbations can cause a CNN to make incorrect predictions. Here we provide insight into this brittleness by investigating the representations of models that are either robust or not robust to image perturbations. Theory suggests that the robustness of a system to these perturbations could be related to the power law exponent of the eigenspectrum of its set of neural responses, where power law exponents closer to and larger than one would indicate a system that is less susceptible to input perturbations. We show that neural responses in mouse and macaque primary visual cortex (V1) obey the predictions of this theory, where their eigenspectra have power law exponents of at least one. We also find that the eigenspectra of model representations decay slowly relative to those observed in neurophysiology and that robust models have eigenspectra that decay slightly faster and have higher power law exponents than those of non-robust models. The slow decay of the eigenspectra suggests that substantial variance in the model responses is related to the encoding of fine stimulus features. We therefore investigated the spatial frequency tuning of artificial neurons and found that a large proportion of them preferred high spatial frequencies and that robust models had preferred spatial frequency distributions more aligned with the measured spatial frequency distribution of macaque V1 cells. Furthermore, robust models were quantitatively better models of V1 than non-robust models. Our results are consistent with other findings that there is a misalignment between human and machine perception. They also suggest that it may be useful to penalize slow-decaying eigenspectra or to bias models to extract features of lower spatial frequencies during task-optimization in order to improve robustness and V1 neural response predictivity.

Dataset Information

Columnar architecture improves noise robustness in a model cortical network.

Publications

Columnar architecture improves noise robustness in a model cortical network.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure