Investigating The Role of Noise Correlations in Learning

Cognitive and Molecular Challenges to Statistical Inference Across Healthy Aging

A fundamental problem in neuroscience is how the brain computes with noisy neurons. An advantage of population codes is that downstream neurons can pool across multiple neurons to reduce the impact of noise. However, this benefit depends on the noise associated with each neuron being independent. Noise correlations refer to the covariance of noise between pairs of neurons, and such correlations can limit the advantages gained from pooling across large neural populations. Indeed, a large body of theoretical work argues that positive noise correlations between similarly tuned neurons reduce the representational capacity of neural populations and are thus detrimental to neural computation. Despite this apparent disadvantage, such noise correlations are observed across many different brain regions, persist even in well-trained subjects, and are dynamically altered in complex tasks. The investigators have advanced the hypothesis that noise correlations may be a neural mechanism for reducing the dimensionality of learning problems. The viability of this hypothesis has been demonstrated in neural network simulations where noise correlations, when embedded in populations with fixed signal-to-noise ratio, enhance the speed and robustness of learning. Here the investigators aim to empirically test this hypothesis, using a combination of computational modeling, fMRI and pupillometry. Establishing a link between noise correlations and learning would open the door to an investigation into how brains navigate a tradeoff between representational capacity and the speed of learning.

Study Overview

• Status: Completed
• Conditions: Noise Correlations; Learning Quality
• Intervention / Treatment: Behavioral: Dynamic perceptual discrimination task; Diagnostic test: fMRI

Detailed Description

Mammalian brains represent information using distributed population codes which provide a number of advantages from robustness to high representational capacity. However, for downstream readout neurons such codes pose formidable high-dimensional learning problems as a very large number of synaptic connections must be adjusted during learning in search of a suitable readout. Our recent theoretical work hypothesized that these high-dimensional learning problems can be simplified by inductive biases implemented through stimulus-independent noise correlations which express the degree to which a pair of neurons covary in their trial-to-trial fluctuations. While noise correlations have traditionally been viewed as providing constraints on representational capacity our recent work demonstrates that they simultaneously constrain readout learning. In some biologically relevant cases, they could theoretically speed learning by shaping the geometry of the underlying neural space to focus the gradient of learning onto task-relevant dimensions. However, this hypothesized role of noise correlations in shaping learning has not yet been empirically tested. Here the investigators elaborate an experimental framework to test the predicted role of noise correlations, as measured through covariation in fMRI multi-voxel BOLD activity patterns for a given stimulus, on learning in both familiar and novel contexts. In familiar contexts, useful noise correlations may be induced by top-down inputs from the prefrontal cortex that signal relevant task dimensions. Thus, the strength of noise correlations in task-relevant dimensions would predict faster learning about task-relevant features. On the other hand, in novel contexts when the relevant task dimensions are unknown, noise correlations may force gradients onto task-irrelevant dimensions and thus impair learning. Therefore, suppressing noise correlations, which might be achieved through neuromodulatory signaling, may speed learning by reducing bias early during learning or after a change in the task-relevant stimulus. Across our Aims, the investigators develop a plan to test the most basic predictions of our computational model using fMRI to characterize the geometry of noise correlations and pupillometry as a proxy for neuromodulatory signaling in human subjects. The planned research will provide the first empirical test of the role of noise correlations in learning.

• Study Type: Interventional
• Enrollment (Actual): 47
• Phase: Not Applicable

Contacts and Locations

Study Locations

• United States
  • Rhode Island
    • Providence, Rhode Island, United States, 02906
      • Brown University

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: Yes

Description

Inclusion Criteria:

• Age above 18
• Normal or correctable vision

Exclusion Criteria:

• Age under 18
• Claustrophobia
• Color blindness
• Neuroleptics medications
• History of drug abuse and/or alcoholism
• Conditions contraindicated for MRI such as:
• Surgical implant that is not MRI compatible
• Metal fragments in the body
• Tattoo with metallic ink
• Eye diseases / impairment:
• Cataracts
• Macular degeneration
• Retinopathies
• Partial vision loss
• Medical history:
• Stroke
• Traumatic brain injury
• Epilepsy
• Schizophrenia
• Manic depression with symptoms including but not limited to psychosis, mania, delusional thinking, and audio/visual hallucinations.

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Basic Science
• Allocation: N/A
• Interventional Model: Single Group Assignment
• Masking: None (Open Label)

Number of Arms

1

Arms and Interventions

Participant Group / Arm

Intervention / Treatment

Experimental: Dynamic perceptual discrimination task; The task featured two task conditions, each of which required the integration of information from both stimulus dimensions. In each condition, participants viewed a stimulus containing motion and color information and were required to specify one of two possible responses. Within each condition, rules and the response mapping changed occasionally, but always by changing on a fixed feature dimension (ie. rightward/purple, leftward/orange). These uncued intra-dimensional shifts involved translational shifts in the learning boundary, requiring them to adapt their decision making within a familiar dimension. These shifts compelled participants to continuously adjust their learning strategies by focusing on the most relevant feature dimension.

Behavioral: Dynamic perceptual discrimination task; The study featured two task conditions, each of which required the integration of information from both stimulus dimensions. In each condition, participants viewed a stimulus containing motion and color information and were required to specify one of two possible responses. Within each condition, rules and the response mapping changed occasionally, but always by changing on a fixed feature dimension (ie. rightward/purple, leftward/orange). These uncued intra-dimensional shifts involved translational shifts in the learning boundary, requiring them to adapt their decision making within a familiar dimension. These shifts compelled participants to continuously adjust their learning strategies by focusing on the most relevant feature dimension.; Diagnostic test: fMRI; Participant brain imaging data will be collected concurrently while performing the perceptual discrimination task.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Participant learning asymmetry in the behavioral task	The participants' learning asymmetries on the task-relevant and task-irrelevant dimensions are evaluated with our reinforcement learning model that recover their learning gradient on the respective learning dimensions.	From the end of the behavioral session to the beginning of the scanning session, typically within a week
Noise correlations in the brain	The investigators will identify the noise correlation - the ratio of trial-by-trial variability associated with a stimulus along the task-relevant versus task-irrelevant coding axes. The coding axes will be decoded from region of interests using multi-voxel patterns associate with individual trials.	Through completion of analysis, an average of 6 months

Collaborators and Investigators

• Sponsor: Brown University
• Collaborators: National Institute of General Medical Sciences (NIGMS)
• Investigators: Principal Investigator: Matthew Nassar, PhD, Brown University

Publications and helpful links

General Publications

• Fusi S, Miller EK, Rigotti M. Why neurons mix: high dimensionality for higher cognition. Curr Opin Neurobiol. 2016 Apr;37:66-74. doi: 10.1016/j.conb.2016.01.010. Epub 2016 Feb 4.
• Cohen MR, Newsome WT. Context-dependent changes in functional circuitry in visual area MT. Neuron. 2008 Oct 9;60(1):162-73. doi: 10.1016/j.neuron.2008.08.007.
• Nassar MR, Scott D, Bhandari A. Noise Correlations for Faster and More Robust Learning. J Neurosci. 2021 Aug 4;41(31):6740-6752. doi: 10.1523/JNEUROSCI.3045-20.2021. Epub 2021 Jun 30.

Study record dates

Study Major Dates

• Study Start (Actual): December 14, 2024
• Primary Completion (Actual): September 1, 2025
• Study Completion (Actual): September 1, 2025

Study Registration Dates

• First Submitted: October 18, 2024
• First Submitted That Met QC Criteria: November 1, 2024
• First Posted (Actual): November 4, 2024

Study Record Updates

• Last Update Posted (Actual): May 26, 2026
• Last Update Submitted That Met QC Criteria: May 21, 2026
• Last Verified: November 1, 2024

More Information

Terms related to this study

• Keywords: neural network; perceptual learning; noise correlations; feature dimensions
• Additional Relevant MeSH Terms: Diagnostic Techniques and Procedures; Diagnosis; Tomography; Diagnostic Imaging; Magnetic Resonance Imaging
• Other Study ID Numbers: 1806002126; 1P30GM149405-01 (U.S. NIH Grant/Contract)

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO
• IPD Plan Description: Anyone interested in IPD should reach out to the Principal Investigator

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: Yes
• product manufactured in and exported from the U.S.: No