Uncertainty Quantification in Working Memory via Moment Neural Networks

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ma, Hengyuan, Lu, Wenlian, Feng, Jianfeng
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916490921377792
author Ma, Hengyuan
Lu, Wenlian
Feng, Jianfeng
author_facet Ma, Hengyuan
Lu, Wenlian
Feng, Jianfeng
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2411_14196
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Uncertainty Quantification in Working Memory via Moment Neural Networks
Ma, Hengyuan
Lu, Wenlian
Feng, Jianfeng
Biological Physics
Neural and Evolutionary Computing
Applications
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.
title Uncertainty Quantification in Working Memory via Moment Neural Networks
topic Biological Physics
Neural and Evolutionary Computing
Applications
url https://arxiv.org/abs/2411.14196