Uncertainty Quantification in Working Memory via Moment Neural Networks

TL;DR

This paper uses moment neural networks to model how the brain quantifies uncertainty in working memory, revealing neural mechanisms and demonstrating comparable performance to humans in uncertainty coding.

Contribution

It introduces a novel application of moment neural networks to neural uncertainty quantification in working memory, linking probabilistic coding with biological neural dynamics.

Findings

MNNs identify firing covariance as a key uncertainty indicator

Model achieves human-like coding precision and uncertainty quantification

Noise and heterogeneity improve working memory performance

Abstract

Humans possess a finely tuned sense of uncertainty that helps anticipate potential errors, vital for adaptive behavior and survival. However, the underlying neural mechanisms remain unclear. This study applies moment neural networks (MNNs) to explore the neural mechanism of uncertainty quantification in working memory (WM). The MNN captures nonlinear coupling of the first two moments in spiking neural networks (SNNs), identifying firing covariance as a key indicator of uncertainty in encoded information. Trained on a WM task, the model demonstrates coding precision and uncertainty quantification comparable to human performance. Analysis reveals a link between the probabilistic and sampling-based coding for uncertainty representation. Transferring the MNN's weights to an SNN replicates these results. Furthermore, the study provides testable predictions demonstrating how noise and heterogeneity enhance WM performance, highlighting their beneficial role rather than being mere biological byproducts. These findings offer insights into how the brain effectively manages uncertainty with exceptional accuracy.