Neural Computations by Asymmetric Networks with Nonlinearities

Author(s):  
Naohiro Ishii ◽  
Toshinori Deguchi ◽  
Masashi Kawaguchi
Author(s):  
Alireza Soltani ◽  
Etienne Koechlin

AbstractThe real world is uncertain, and while ever changing, it constantly presents itself in terms of new sets of behavioral options. To attain the flexibility required to tackle these challenges successfully, most mammalian brains are equipped with certain computational abilities that rely on the prefrontal cortex (PFC). By examining learning in terms of internal models associating stimuli, actions, and outcomes, we argue here that adaptive behavior relies on specific interactions between multiple systems including: (1) selective models learning stimulus–action associations through rewards; (2) predictive models learning stimulus- and/or action–outcome associations through statistical inferences anticipating behavioral outcomes; and (3) contextual models learning external cues associated with latent states of the environment. Critically, the PFC combines these internal models by forming task sets to drive behavior and, moreover, constantly evaluates the reliability of actor task sets in predicting external contingencies to switch between task sets or create new ones. We review different models of adaptive behavior to demonstrate how their components map onto this unifying framework and specific PFC regions. Finally, we discuss how our framework may help to better understand the neural computations and the cognitive architecture of PFC regions guiding adaptive behavior.


2017 ◽  
Vol 13 (8) ◽  
pp. e1005723 ◽  
Author(s):  
Vickie Li ◽  
Santiago Herce Castañón ◽  
Joshua A. Solomon ◽  
Hildward Vandormael ◽  
Christopher Summerfield
Keyword(s):  

2021 ◽  
pp. 1-44
Author(s):  
Edoardo Gallo ◽  
Chang Yan

Abstract The tension between efficiency and equilibrium is a central feature of economic systems. We examine this trade-off in a network game with a unique Nash equilibrium in which agents can achieve a higher payoff by following a “collaborative norm”. Subjects establish and maintain a collaborative norm in the circle, but the norm weakens with the introduction of one hub connected to everyone in the wheel. In complex and asymmetric networks of 15 and 21 nodes, the norm disappears and subjects’ play converges to Nash. We provide evidence that subjects base their decisions on their degree, rather than the overall network structure.


Author(s):  
Johannes Mehrer ◽  
Courtney J. Spoerer ◽  
Nikolaus Kriegeskorte ◽  
Tim C. Kietzmann

AbstractDeep neural networks (DNNs) excel at visual recognition tasks and are increasingly used as a modelling framework for neural computations in the primate brain. However, each DNN instance, just like each individual brain, has a unique connectivity and representational profile. Here, we investigate individual differences among DNN instances that arise from varying only the random initialization of the network weights. Using representational similarity analysis, we demonstrate that this minimal change in initial conditions prior to training leads to substantial differences in intermediate and higher-level network representations, despite achieving indistinguishable network-level classification performance. We locate the origins of the effects in an under-constrained alignment of category exemplars, rather than a misalignment of category centroids. Furthermore, while network regularization can increase the consistency of learned representations, considerable differences remain. These results suggest that computational neuroscientists working with DNNs should base their inferences on multiple networks instances instead of single off-the-shelf networks.


2017 ◽  
Author(s):  
William J Harrison ◽  
Reuben Rideaux

ABSTRACTThe extent to which visual inference is shaped by attentional goals is unclear. Voluntary attention may simply modulate the priority with which information is accessed by higher cognitive functions involved in perceptual decision making. Alternatively, voluntary attention may influence fundamental visual processes, such as those involved in segmenting an incoming retinal signal into a structured scene of coherent objects, thereby determining perceptual organisation. Here we tested whether the segmentation and integration of visual form can be determined by an observer’s goals by exploiting a novel variant of the classical Kanizsa figure. We generated predictions about the influence of attention with a machine classifier, and tested these predictions with a psychophysical response classification technique. Despite seeing the same image on each trial, observers’ perception of illusory spatial structure depended on their attentional goals. These attention-contingent illusory contours directly conflicted with equally plausible visual form implied by the geometry of the stimulus, revealing that attentional selection can determine the perceived layout of a fragmented scene. Attentional goals, therefore, not only select pre-computed features or regions of space for prioritised processing, but, under certain conditions, also greatly influence perceptual organisation and thus visual appearance.SIGNIFICANCE STATEMENTThe extent to which higher cognitive functions can influence perceptual organisation is debated. The role of voluntary spatial attention, the ability to focus on only some parts of a scene, has been particularly controversial among neuroscientists and psychologists who aim to uncover the basic neural computations involved in grouping image features into coherent objects. To address this issue, we repeatedly presented the same novel ambiguous image to observers and changed their attentional goals by having them make fine spatial judgements about only some elements of the image. We found that observers’ attentional goals determine the perceived organisation of multiple illusory shapes. We thus reveal that voluntary spatial attention can control the fundamental processes that determine perceptual organisation.


2020 ◽  
Author(s):  
Adrienne C. Loewke ◽  
Adelaide R. Minerva ◽  
Alexandra B. Nelson ◽  
Anatol C. Kreitzer ◽  
Lisa A. Gunaydin

ABSTRACTThe dorsomedial prefrontal cortex (dmPFC) has been linked to approach-avoidance behavior and decision-making under conflict, key neural computations thought to be altered in anxiety disorders. However, the heterogeneity of efferent prefrontal projections has obscured identification of the specific top-down neural pathways regulating these anxiety-related behaviors. While the dmPFC-amygdala circuit has long been implicated in controlling reflexive fear responses, recent work suggests that this circuit is less important for avoidance behavior. We hypothesized that dmPFC neurons projecting to the dorsomedial striatum (DMS) represent a subset of prefrontal neurons that robustly encode and drive approach-avoidance behavior. Using fiber photometry recording during the elevated zero maze (EZM) task, we show heightened neural activity in prefrontal and fronto-striatal projection neurons, but not fronto-amydalar projection neurons, during exploration of the anxiogenic open arms of the maze. Additionally, through pathway-specific optogenetics we demonstrate that this fronto-striatal projection preferentially excites postsynaptic D1 receptor-expressing medium spiny neurons in the DMS and bidirectionally controls avoidance behavior. We conclude that this striatal-projecting subpopulation of prefrontal neurons regulates approach-avoidance conflict, supporting a model for prefrontal control of defensive behavior in which the dmPFC-amygdala projection controls reflexive fear behavior and the dmPFC-striatum projection controls anxious avoidance behavior. Our findings identify this fronto-striatal circuit as a valuable therapeutic target for developing interventions to alleviate excessive avoidance behavior in anxiety disorders.


Sign in / Sign up

Export Citation Format

Share Document