Functional brain network reconfiguration during learning in a dynamic environment

Abstract

<jats:title>Abstract</jats:title><jats:p>When learning about dynamic and uncertain environments, people should update their beliefs most strongly when new evidence is most informative, such as when the environment undergoes a surprising change or existing beliefs are highly uncertain. Here we show that modulations of surprise and uncertainty are encoded in a particular, temporally dynamic pattern of whole-brain functional connectivity, and this encoding is enhanced in individuals that adapt their learning dynamics more appropriately in response to these factors. The key feature of this whole-brain pattern of functional connectivity is stronger connectivity, or functional integration, between the fronto-parietal and other functional systems. Our results provide new insights regarding the association between dynamic adjustments in learning and dynamic, large-scale changes in functional connectivity across the brain.</jats:p>

The stability flexibility tradeoff and the dark side of detail

“…Dynamic fluctuations in learning rate relate to overall arousal levels as measured by pupil diameter (Nassar et al, 2012), as well as activation in a network that includes insula, dorsomedial prefrontal cortex, and parietal cortex, and parts of dorsolateral prefrontal cortex (Behrens et al, 2007;McGuire et al, 2014;Payzan-LeNestour, Dunne, Bossaerts, & O'Doherty, 2013). Functional connectivity over a subgraph that includes many of these regions, and is closely related to both the salience and central executive networks described above, predicts individual differences in learning behavior (Kao et al, 2019). Given the well established connectivity differences in ASD, it is possible both attention to detail and adaptive learning in our task are jointly driven by individual differences in functional connectivity, and we hope that our work motivates future explorations of these brain-behavior relationships.…”

Adaptive Learning through Temporal Dynamics of State Representation

“…Such probabilities can be inferred using a Bayesian learning model calibrated to the environmental structure; however, we show that they could also be estimated from the output layer of our network itself. Previous work has suggested that changepoint and oddball probability are reflected by BOLD activations in both cortical and subcortical regions (Yu and Dayan, 2005;Nassar et al, 2012;O'Reilly et al, 2013;McGuire et al, 2014;D'Acremont and Bossaerts, 2016;Meyniel and Dehaene, 2017;Nassar et al, 2019a;Kao et al, 2020). While such signals have previously been interpreted as early-stage computations performed in the service of computing a learning rate, our work suggests that they serve another purpose, namely in signaling the need to change the active context representation.…”

Locating causal hubs of memory consolidation in spontaneous brain network

“…Based on resting-state functional magnetic resonance imaging (rsfMRI), functional connectivity (FC) measured as the statistical relationship between regional hemodynamic activities has been found to form large-scale resting-state networks (RSNs) that associate with cognition and its age-related disorders at the system level (Biswal et al, 1995; Craddock et al, 2013; Mill et al, 2017). For instance, FC correlates with and predicts memory performance and can be reshaped by learning, indicating strong behavioral relevance of the RSNs (Albert et al, 2009; Bassett et al, 2011; Kao et al, 2020; Tambini et al, 2010). The disruption of RSNs in association with cognitive impairment and neuropathology provide further evidence for the involvement of these networks in the etiology and progression in diseases (Greicius et al, 2003; Pievani et al, 2014; Yu et al, 2021).…”