Scientific Understanding of Consciousness
Consciousness as an Emergent Property of Thalamocortical Activity

Neural Network — Autoassociation Attractor Systems

 

 

Autoassociative memories, or attractor neural networks, store memories, each one of which is represented by a pattern of neural activity. (Rolls & Deco; Noisy Brain, 31)

Autoassociative networks can recall the appropriate memory from the network when provided with a fragment of one of the memories. This is called completion. (Rolls & Deco; Noisy Brain, 31)

Any two cells or systems of cells that are repeatedly active at the same time will tend to become "associated," so that activity in one facilitates activity in the other. (Hebb; Organization of Behavior, 70)

 

Research Study — Chained Activation of Neuronal Assemblies supports Major Cognitive Processes

Link to — Associative nature of Memory

 

A feature of autoassociative memory is that it is content addressable; i.e. the information in the memory can be accessed if just the contents of the memory (or in part of the contents of the memory) are used. (Rolls & Deco; Noisy Brain, 31)

Autoassociation memory can be used as a short-term memory, in which iterative processing round the recurrent collateral connection loop keeps a representation active by continuing neuronal firing. (Rolls & Deco; Noisy Brain, 31)

The short-term memory reflected in continuing neuronal firing for several hundred milliseconds after a visual stimulus is removed, which is present in visual cortical areas such as inferior temporal visual cortex, is probably implemented in this way. (Rolls & Deco; Noisy Brain, 31)

Autoassociation memory appears to be used in a short-term memory in the prefrontal cortex.  Temporal visual cortical areas have connections to the ventrolateral prefrontal cortex that help to implement short-term memory for visual stimuli. (Rolls & Deco; Noisy Brain, 31)

Attractor networks are implemented by the forward and backward connections between cortical areas. (Rolls & Deco; Noisy Brain, 31)

An autoassociation network with the rapid synaptic plasticity can learn each memory in one trial.  Because of its one-shot rapid learning, and ability to complete, this type of network is well suited for episodic memory storage, in which each past episode must be stored and recalled later from fragments, and kept separate from other episodic memories. (Rolls & Deco; Noisy Brain, 31)

Perhaps the most important and useful property of autoassociative memories is that they complete an incomplete input vector, allowing recall of a whole memory from a small fraction of it. (Rolls & Deco; Noisy Brain, 35)

One factor in the rapid dynamics of associative networks with brain-like 'integrate-and-fire' membrane and synaptic properties is that with ongoing spontaneous activity, some of the neurons in the network are close to threshold already, before the recall cue is applied, and hence some of the neurons are very quickly pushed by the recall cue into firing, perhaps within 1-2 ms. (Rolls & Deco; Noisy Brain, 36)

Autoassociation attractor systems have two types of stable fixed points: (1) a spontaneous state with a low firing rate, and (2) one or more attractor states in which the positive feedback implemented by the recurrent collateral connections maintain a high firing rate. (Rolls & Deco; Noisy Brain, 73)

The high firing rate attractor state is sometimes referred to as the persistent state, because the high firing rate persists to maintain a set of neurons active, which might implement a short-term memory. (Rolls & Deco; Noisy Brain, 73)

The stable points in the system can be visualized in an energy landscape. (Rolls & Deco; Noisy Brain, 73)

The area in the energy landscape within which the system will move to a stable attractor state is call the basin of attraction. (Rolls & Deco; Noisy Brain, 73)

The attractor dynamics can be pictured by energy landscapes, which indicate the basins of attraction by valleys, and the attractor states or fixed points by the bottom of the valleys. (Rolls & Deco; Noisy Brain, 74)

The stability of an attractor is characterized by the average time in which the system stays in the basin of attraction under the influence of noise. (Rolls & Deco; Noisy Brain, 74)

 

An autoassociation network with the rapid synaptic plasticity can learn each memory in one trial.  Because of its one-shot rapid learning, and ability to complete, this type of network is well suited for episodic memory storage, in which each past episode must be stored and recalled later from fragments, and kept separate from other episodic memories. (Rolls & Deco; Noisy Brain, 31)

Consciousness depends upon dialog between diverse regions of the brain, associated with independent psychological functions such as perception, emotion, memory, and action. (Zeman; Consciousness, 291)  [thalamocortical system]  [Edelman's dynamic core]

Limited controlled synchronization of rapid neuronal discharge might play a role in perception, memory and movement. (Zeman; Consciousness, 293)

Autoassociative memories, or attractor neural networks, store memories each one of which is represented by a pattern of neural activity. (Rolls & Deco; Noisy Brain, 31)

Memories are stored in the recurrent synaptic connections between the neurons of the network, for example in the recurrent collateral connections between cortical pyramidal cells. (Rolls & Deco; Noisy Brain, 31)

Autoassociative networks can recall the appropriate memory from the network when provided with a fragment of one of the memories. This is called completion. (Rolls & Deco; Noisy Brain, 31)

Autoassociation memory appears to be used in a short-term memory in the prefrontal cortex. (Rolls & Deco; Noisy Brain, 31)

Pattern theory

Pattern theory says that tightly coupled cortical areas seek, via some kind of relaxation functionality, to arrive at a mutual agreement in which lower areas' specific data form a fit with known, more abstract categorizations stored in higher areas' memory of prior activity. (Mumford; Neuronal Architectures, 135)

Bottom-up assertions of facts have to be included along with top-down memories of expected patterns. (Mumford; Neuronal Architectures, 135)

In hippocampus, it has been estimated that a given axon usually makes only a single synapse (average estimated to be 1.3) on its target cell. (Stevens; Cortical Theory, 241)

Lateral geniculate axons that project to visual cortex make only one or a few (up to about eight) synapses on their targets. (Stevens; Cortical Theory, 241)

 

 

Return to — Neural Network