Figure 1 - available via license: Creative Commons Attribution-NoDerivatives 4.0 International
Content may be subject to copyright.
Network Schematics and Sample Trial. I: input; C choice layer synaptic gating variable; T : intermediate transform layer synaptic gating variable; red: choice A; blue: choice B; black: attribute 1; white: 2; arrows: excitation; circles: inhibition. Both networks include an input layer and a final choice layer. (A) The Linear Network consists of two layers, with attribute signals from the input layer directly transmitted to the final choice layer. (B) The Hierarchical Network includes an additional intermediate layer that performs a functional transformation on the attribute signals prior to their passing to the choice area. (C) A sample trial of the Hierarchical Network, with intermediate layer weights J+ = 0.34 nA and J− = −0.02. For each area, the X-axis indicates time, while the Y-axis indicates the population firing rate. The first vertical line indicates the onset of the offer value signal, and the second vertical line indicates termination of the offer value signal.
Source publication
We investigated two-attribute, two-alternative decision-making in a hierarchical neural network with three layers: an input layer encoding choice alternative attribute values; an intermediate layer of modules processing separate attributes; and a choice layer producing the decision. Depending on intermediate layer excitatory-inhibitory (E/I) tone,...
Contexts in source publication
Context 1
... the name suggests, the Linear Network computed an exact linear weighted sum of the attributes (weight = 0.5). Thus, with the Linear Network, the choice area received a signal that was linearly translated from the presented attributes of the choice alternatives ( Figure 1A). This network has a similar architecture to that presented in (Rustichini and Padoa-Schioppa 2015), and represents the case where all attribute signals are fully available to a choice area. ...
Context 2
... the Hierarchical Network ( Figure 1B), inputs were first transformed by intermediate, attribute-specific areas. The intermediate layer transforming the input signal was a nonlinear dynamical system, composed of attractor states associated with each choice alternative (see Methods). ...
Context 3
... continuous output signals were then fed into a choice area that determined the decision on a given trial. A single trial of the Hierarchical Network is shown in Figure 1C. This framework, similar in structure to that presented in (Hunt, Dolan, and Behrens 2014), represents the simplest form of a network with parallel processing streams, where transformations are performed on attribute signals passing through specialized areas prior to the final decision. ...
Context 4
... verified the networks were able to perform the task, we next examined how they utilize attribute information. While the Linear Network performs a simple linear sum of the attributes, the Hierarchical Network first passes the input signals through attribute-specific areas ( Figure 1B and Figure 3A). This transform of inputs is defined by the recurrent excitation and cross-inhibition of areas in the intermediate layer. ...
Context 5
... will consider and compare two models of multi-attribute decisions. In the linear model (LM), attribute input values (I) for each choice alternative were linearly combined to reach a choice value C A or C B , and then the maximum of those values was taken as the choice (as in Figure S1A). The sequential max model (SM) consisted of a series of max operations (as in Figure S1B-C). ...
Context 6
... the linear model (LM), attribute input values (I) for each choice alternative were linearly combined to reach a choice value C A or C B , and then the maximum of those values was taken as the choice (as in Figure S1A). The sequential max model (SM) consisted of a series of max operations (as in Figure S1B-C). These two models are extreme cases of decision regimes adopted by the Hierarchical Network, as described in the main body of the paper. ...
Context 7
... LM was meant to capture the Linear Network, along with the case of the Hierarchical Network model performing a linear addition of the attribute values. The operations of the LM are shown in Figure S1A. In this model, the attributes associated with each choice alternative are first summed, such that, C A = I A,1 + I A,2 , and C B = I B,1 + I B,2 . ...
Context 8
... SM is named for the series of max operations performed, first at the input level, and then with the final decision. Those operations are shown in Figure S1B-C. This model produces the concave decision behavior described in the main text. ...
Context 9
... decision process is shown in on the left of Figure S1B. Since only the winners of the first max ...
Context 10
... copyright holder for this . http://dx.doi.org/10.1101/2020.01.28.923649 doi: bioRxiv preprint first posted online Jan. 29, 2020; operation (bolded in Figure S1) are passed to the next stage, the choice values will be, C A = I A,1 + I A,2 , and C B = 0. ...
Similar publications
Single-cell and bulk genomics assays have complementary strengths and weaknesses, and alone neither strategy can fully capture regulatory elements across the diversity of cells in complex tissues. We present CellWalker, a method that integrates single-cell open chromatin (scATAC-seq) data with gene expression (RNA-seq) and other data types using a...
Citations
... Robust inhibition and long inhibitory time constants should contribute to the extension of the time window for signal summation and thus extend local temporal receptive fields. In artificial hierarchical networks, environmental uncertainty can be dynamically captured by variations of the E/I tone (Pettine, Louie, Murray, & Wang, 2020). In humans, dynamical integration of environmental uncertainty is circumscribed to the MCC (Behrens et al., 2007). ...
The ability to integrate information across time at multiple timescales is a vital element of adaptive behavior, because it provides the capacity to link events separated in time, extract useful information from previous events and actions, and to construct plans for behavior over time. Here we make the argument that this information integration capacity is a central function of the midcingulate cortex (MCC), by reviewing the anatomical, intrinsic network, neurophysiological, and behavioral properties of MCC. The MCC is the region of the medial wall situated dorsal to the corpus callosum and sometimes referred to as dACC. It is positioned within the densely connected core network of the primate brain, with a rich diversity of cognitive, somatomotor and autonomic connections. Furthermore, the MCC shows strong local network inhibition which appears to control the metastability of the region—an established feature of many cortical networks in which the neural dynamics move through a series of quasi-stationary states. We propose that the strong local inhibition in MCC leads to particularly long dynamic state durations, and so less frequent transitions. Apparently as a result of these anatomical features and synaptic and ionic determinants, the MCC cells display the longest neuronal timescales among a range of recorded cortical areas. We conclude that the anatomical position, intrinsic properties, and local network interactions of MCC make it a uniquely positioned cortical area to perform the integration of diverse information over time that is necessary for behavioral adaptation.
... Robust inhibition and long inhibitory time constants should contribute to the extension of the time window for signal summation and thus extend local temporal receptive fields. In artificial hierarchical networks, environmental uncertainty can be dynamically captured by variations of the E/I tone (Pettine, Louie, Murray, & Wang, 2020). In humans, dynamical integration of environmental uncertainty is circumscribed to the MCC (Behrens et al., 2007). ...
