Image memorability is predicted by discriminability and similarity in different stages of a convolutional neural network.

The features of an image can be represented at multiple levels-from its low-level visual properties to high-level meaning. What drives some images to be memorable while others are forgettable? We address this question across two behavioral experiments. In the first, different layers of a convolutional neural network (CNN), which represent progressively higher levels of features, were used to select the images that would be shown to 100 participants through a form of prospective assignment. Here, the discriminability/similarity of an image with others, according to different CNN layers dictated the images presented to different groups, who made a simple indoor versus outdoor judgment for each scene. We found that participants remember more scene images that were selected based on their low-level discriminability or high-level similarity. A second experiment replicated these results in an independent sample of 50 participants, with a different order of postencoding tasks. Together, these experiments provide evidence that both discriminability and similarity, at different visual levels, predict image memorability.

Neural correlates of mindful disengagement from worry.

Uncontrollable worry is a hallmark of generalized anxiety disorder and a transdiagnostic feature of psychopathology. Mindfulness-based strategies show promise for treating worry, but it is unknown which specific strategies are most beneficial, and how these skills might operate on a neurobiological level. We recruited 40 participants with clinically significant worry to undergo functional magnetic resonance imaging while engaging in real-time, idiographic worry and instructed disengagement using two mindfulness strategies (focused attention, acceptance) and one comparison strategy (suppression). Hypotheses were preregistered and partially supported. All disengagement strategies downregulated default mode and upregulated frontoparietal and salience networks, suggesting some shared mechanisms. Focused attention was most effective for promoting disengagement and elicited decreased activity in cognitive control and sensorimotor regions. Successful disengagement was associated with increased activity in rostrolateral prefrontal cortex and functional connectivity between posterior cingulate and primary somatosensory cortex. Findings support the role of cognitive control and somatosensory networks in disengagement from worry and suggest common and distinct mechanisms of disengagement, with focused attention a particularly promising strategy. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Novel approaches to functional lateralization: Assessing information in activity patterns across hemispheres and more accurately identifying structural homologues.

Functional lateralization is typically measured by comparing activation levels across the right and left hemispheres of the brain. Significant additional information, however, exists within distributed multi-voxel patterns of activity - a format not detectable by traditional activation-based analysis of functional magnetic resonance imaging (fMRI) data. We introduce and test two methods -one anatomical, one functional- that allow hemispheric information asymmetries to be detected. We first introduce and apply a novel tool that draws on brain 'surface fingerprints' to pair every location in one hemisphere with its hemispheric homologue. We use anatomical data to show that this approach is more accurate than the common distance-from-midline method for comparing bilateral regions. Next, we introduce a complementary analysis method that quantifies multivariate laterality in functional data. This new 'multivariate Laterality Index' (mLI) reflects both quantitative and qualitative information-differences across homologous activity patterns. We apply the technique here to functional data collected as participants viewed faces and non-faces. Using the previously generated surface fingerprints to pair-up homologous searchlights in each hemisphere, we use the novel multivariate laterality technique to identify face-information asymmetries across right and left counterparts of the fusiform gyrus, inferior temporal gyrus, superior parietal lobule, and early visual areas. The typical location of the fusiform face area has greater information asymmetry for faces than for shapes. More generally, we argue that the field should consider an information-based approach to lateralization.