Spectral-Stimulus Information for Self-Supervised Stimulus Encoding

TL;DR

This paper introduces new correlation-aware information measures to analyze neural population encoding in spatial navigation, revealing how neurons collectively encode stimuli and enabling the training of RNNs to develop place and head direction cells.

Contribution

It presents novel spectral-stimulus information measures that account for neural correlations, advancing understanding of population encoding and enabling self-supervised learning of spatial cells in neural networks.

Findings

Spectral-stimulus information is maximized with localized, non-overlapping firing fields.

Neural populations encode stimuli more efficiently when considering correlations.

Self-supervised training of RNNs leads to emergence of place and head direction cells.

Abstract

Mammalian spatial navigation relies on specialized neurons, such as place and grid cells, which encode position based on self-motion and environmental cues. While extensive research has explored the computational role of grid cells, the principles underlying efficient place cell coding remain less understood. Existing spatial information rate measures primarily assess single-neuron encoding, limiting insights into population-level representations, while, the role of correlation in neural coding remains a subject of considerable debate. To address this, we introduce novel, correlation-aware information-theoretic measures that quantify the encoding efficiency of multiple neurons, including the joint stimulus information rate for neuron pairs and the spectral-stimulus information for arbitrary sized populations. The spectral-stimulus information, defined as the leading eigenvalue of the stimulus information matrix, is maximized when neurons exhibit localized, non-overlapping firing fields, mirroring place cell and head direction cell activity. We apply these measures to neural data recorded in mice and monkeys, elucidating differences in encoding efficiency across neuronal pairs and populations. Then, we demonstrate that these measures can be used to train recurrent neural networks (RNNs) via self-supervised learning, leading to the emergence of place cells and head direction cells. Our findings highlight how neural populations collectively encode stimuli, offering a more comprehensive framework for understanding stimulus encoding and optimizing artificial navigation systems in novel environments.