Limits...
The necessity of connection structures in neural models of variable binding.

van der Velde F, de Kamps M - Cogn Neurodyn (2015)

Bottom Line: Contrary to his conclusion, connectivity based models are the only viable candidates for models of novel variable binding because they are the only type of models that can produce behavior.Feldman's analysis that novel variable binding can be achieved without existing connection structures seems to result from analyzing the binding problem in a wrong frame of reference, in particular in an outside instead of the required inside frame of reference.We will illustrate binding with this type of model with episode binding and the binding of words, including novel words, in sentence structures.

View Article: PubMed Central - PubMed

Affiliation: Technical Cognition, CPE-CTIT, University of Twente, P.O. Box 217, Enschede, 7500 AE The Netherlands ; IO, Leiden University, Leiden, The Netherlands.

ABSTRACT
In his review of neural binding problems, Feldman (Cogn Neurodyn 7:1-11, 2013) addressed two types of models as solutions of (novel) variable binding. The one type uses labels such as phase synchrony of activation. The other ('connectivity based') type uses dedicated connections structures to achieve novel variable binding. Feldman argued that label (synchrony) based models are the only possible candidates to handle novel variable binding, whereas connectivity based models lack the flexibility required for that. We argue and illustrate that Feldman's analysis is incorrect. Contrary to his conclusion, connectivity based models are the only viable candidates for models of novel variable binding because they are the only type of models that can produce behavior. We will show that the label (synchrony) based models analyzed by Feldman are in fact examples of connectivity based models. Feldman's analysis that novel variable binding can be achieved without existing connection structures seems to result from analyzing the binding problem in a wrong frame of reference, in particular in an outside instead of the required inside frame of reference. Connectivity based models can be models of novel variable binding when they possess a connection structure that resembles a small-world network, as found in the brain. We will illustrate binding with this type of model with episode binding and the binding of words, including novel words, in sentence structures.

No MeSH data available.


Related in: MedlinePlus

Network structure in episode binding. Two arbitrary items (object, events) A and B, represented in the neocortex, are bound in an episode via connections and rapid long-term potentiation in the hippocampus and surrounding medial temporal lobe (after Norman and O’Reilly 2003). Ovals represent neurons (or populations of neurons). Gray ovals are active. When A is activated in the neocortex, it will activate B through the (temporal) connection structure between them in the hippocampus (and medial temporal lobe)
© Copyright Policy - OpenAccess
Related In: Results  -  Collection


getmorefigures.php?uid=PMC4491338&req=5

Fig4: Network structure in episode binding. Two arbitrary items (object, events) A and B, represented in the neocortex, are bound in an episode via connections and rapid long-term potentiation in the hippocampus and surrounding medial temporal lobe (after Norman and O’Reilly 2003). Ovals represent neurons (or populations of neurons). Gray ovals are active. When A is activated in the neocortex, it will activate B through the (temporal) connection structure between them in the hippocampus (and medial temporal lobe)

Mentions: The combination of (early) LTP and the connection structure resembling a small-world network provide the basis for arbitrary episode binding. Figure 4 schematically illustrates the binding of two arbitrary items (e.g., persons, objects, events) A and B that co-occur in an episode. The neurons representing (processing) the items somewhere in the neocortex are connected to the hippocampus (via the medial temporal cortex). These connections are sparse in the sense that two different items activate substantially different sets of neurons in the hippocampus, which results in pattern separation (Norman and O’Reilly 2003). LTP ensures a (temporal) strengthening of these connections so that both items are temporarily stored in memory. However, connections within the hippocampus (CA3) are also strongly recurrent. With LTP this results in the (temporal) interconnection between the representations of A and B in the hippocampus. As a result, A and B are bound in an episode.Fig. 4


The necessity of connection structures in neural models of variable binding.

van der Velde F, de Kamps M - Cogn Neurodyn (2015)

Network structure in episode binding. Two arbitrary items (object, events) A and B, represented in the neocortex, are bound in an episode via connections and rapid long-term potentiation in the hippocampus and surrounding medial temporal lobe (after Norman and O’Reilly 2003). Ovals represent neurons (or populations of neurons). Gray ovals are active. When A is activated in the neocortex, it will activate B through the (temporal) connection structure between them in the hippocampus (and medial temporal lobe)
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC4491338&req=5

Fig4: Network structure in episode binding. Two arbitrary items (object, events) A and B, represented in the neocortex, are bound in an episode via connections and rapid long-term potentiation in the hippocampus and surrounding medial temporal lobe (after Norman and O’Reilly 2003). Ovals represent neurons (or populations of neurons). Gray ovals are active. When A is activated in the neocortex, it will activate B through the (temporal) connection structure between them in the hippocampus (and medial temporal lobe)
Mentions: The combination of (early) LTP and the connection structure resembling a small-world network provide the basis for arbitrary episode binding. Figure 4 schematically illustrates the binding of two arbitrary items (e.g., persons, objects, events) A and B that co-occur in an episode. The neurons representing (processing) the items somewhere in the neocortex are connected to the hippocampus (via the medial temporal cortex). These connections are sparse in the sense that two different items activate substantially different sets of neurons in the hippocampus, which results in pattern separation (Norman and O’Reilly 2003). LTP ensures a (temporal) strengthening of these connections so that both items are temporarily stored in memory. However, connections within the hippocampus (CA3) are also strongly recurrent. With LTP this results in the (temporal) interconnection between the representations of A and B in the hippocampus. As a result, A and B are bound in an episode.Fig. 4

Bottom Line: Contrary to his conclusion, connectivity based models are the only viable candidates for models of novel variable binding because they are the only type of models that can produce behavior.Feldman's analysis that novel variable binding can be achieved without existing connection structures seems to result from analyzing the binding problem in a wrong frame of reference, in particular in an outside instead of the required inside frame of reference.We will illustrate binding with this type of model with episode binding and the binding of words, including novel words, in sentence structures.

View Article: PubMed Central - PubMed

Affiliation: Technical Cognition, CPE-CTIT, University of Twente, P.O. Box 217, Enschede, 7500 AE The Netherlands ; IO, Leiden University, Leiden, The Netherlands.

ABSTRACT
In his review of neural binding problems, Feldman (Cogn Neurodyn 7:1-11, 2013) addressed two types of models as solutions of (novel) variable binding. The one type uses labels such as phase synchrony of activation. The other ('connectivity based') type uses dedicated connections structures to achieve novel variable binding. Feldman argued that label (synchrony) based models are the only possible candidates to handle novel variable binding, whereas connectivity based models lack the flexibility required for that. We argue and illustrate that Feldman's analysis is incorrect. Contrary to his conclusion, connectivity based models are the only viable candidates for models of novel variable binding because they are the only type of models that can produce behavior. We will show that the label (synchrony) based models analyzed by Feldman are in fact examples of connectivity based models. Feldman's analysis that novel variable binding can be achieved without existing connection structures seems to result from analyzing the binding problem in a wrong frame of reference, in particular in an outside instead of the required inside frame of reference. Connectivity based models can be models of novel variable binding when they possess a connection structure that resembles a small-world network, as found in the brain. We will illustrate binding with this type of model with episode binding and the binding of words, including novel words, in sentence structures.

No MeSH data available.


Related in: MedlinePlus