Neural system identification for large populations separating "what" and "where"

TL;DR

This paper introduces a scalable CNN architecture with a sparse readout layer that effectively separates 'what' and 'where' in neural data, enabling accurate identification of large neural populations with limited data.

Contribution

The authors propose a novel CNN design with a sparse readout layer that factorizes spatial and feature dimensions, improving scalability and accuracy in neural system identification.

Findings

Outperforms existing models in identifying mouse V1 neurons.

Scales efficiently to thousands of neurons with limited data.

End-to-end training of the proposed architecture is feasible.

Abstract

Neuroscientists classify neurons into different types that perform similar computations at different locations in the visual field. Traditional methods for neural system identification do not capitalize on this separation of 'what' and 'where'. Learning deep convolutional feature spaces that are shared among many neurons provides an exciting path forward, but the architectural design needs to account for data limitations: While new experimental techniques enable recordings from thousands of neurons, experimental time is limited so that one can sample only a small fraction of each neuron's response space. Here, we show that a major bottleneck for fitting convolutional neural networks (CNNs) to neural data is the estimation of the individual receptive field locations, a problem that has been scratched only at the surface thus far. We propose a CNN architecture with a sparse readout layer factorizing the spatial (where) and feature (what) dimensions. Our network scales well to thousands of neurons and short recordings and can be trained end-to-end. We evaluate this architecture on ground-truth data to explore the challenges and limitations of CNN-based system identification. Moreover, we show that our network model outperforms current state-of-the art system identification models of mouse primary visual cortex.