An extension of the localist representation theory: grandmother cells are also widely used in the brain

Roy, Asim

doi:10.3389/fpsyg.2013.00300

Cited by 13 publications

(11 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, looking for localist coding of cognitive entities might be a productive scientific strategy, deserving to be pursued (cf. Barlow, 1972;Grainger & Jacobs, 1998;Gross, 2002;Roy, 2012Roy, , 2013Roy, , 2015Roy, , 2017Thorpe, 2011, for other defenses). However, as argued here, one should probably try to localize something like CFDs rather than cognitive functions.…”

Section: Cognitive Functionsmentioning

confidence: 99%

Elements, structure and dynamics of the mind-brain system. A model based on 'Cognitive Feature Detectors'

Dn¹

2020

Preprint

View full text Add to dashboard Cite

General principles of a unified theory of cognition and its relationships with the brain are outlined, integrating insights from several existing theories and approaches. Structure, dynamics, and development of the neural/cognitive system are described. Cognitive Feature Detectors (CFDs) are the fundamental units: they have a (minimalist) semantic value, strictly related to their location in a cognitive space. A cognitivo-neural isomorphism is postulated between CFDs in this cognitive space, and single (or small sets of) neurons in the neural network. The cognitive space is made of two parts (executive and perceptive), densely connected, both hierarchically organized and clustered in cognitive maps. Within maps, inhibitory connections implement competition between incompatible CFDs. This theoretical sketch is hoped to help addressing general theoretical issues such as the relationship between brain and mind, on the basis of the neural localization of cognitive elements.

show abstract

Section: Cognitive Functionsmentioning

confidence: 99%

Elements, structure and dynamics of the mind-brain system. A model based on 'Cognitive Feature Detectors'

Dn¹

2020

Preprint

View full text Add to dashboard Cite

show abstract

“…It seems that the weight of observational and experimental evidence favors the belief that such cells do exist (Gross, 2002; Roy, 2013). This is consistent with the observation that people who have suffered a stroke or are suffering from dementia may lose the ability to recognize members of their close family.…”

Section: Some More Detailmentioning

confidence: 99%

Information Compression, Multiple Alignment, and the Representation and Processing of Knowledge in the Brain

Wolff¹

2016

Front. Psychol.

View full text Add to dashboard Cite

The SP theory of intelligence, with its realization in the SP computer model, aims to simplify and integrate observations and concepts across artificial intelligence, mainstream computing, mathematics, and human perception and cognition, with information compression as a unifying theme. This paper describes how abstract structures and processes in the theory may be realized in terms of neurons, their interconnections, and the transmission of signals between neurons. This part of the SP theory—SP-neural—is a tentative and partial model for the representation and processing of knowledge in the brain. Empirical support for the SP theory—outlined in the paper—provides indirect support for SP-neural. In the abstract part of the SP theory (SP-abstract), all kinds of knowledge are represented with patterns, where a pattern is an array of atomic symbols in one or two dimensions. In SP-neural, the concept of a “pattern” is realized as an array of neurons called a pattern assembly, similar to Hebb's concept of a “cell assembly” but with important differences. Central to the processing of information in SP-abstract is information compression via the matching and unification of patterns (ICMUP) and, more specifically, information compression via the powerful concept of multiple alignment, borrowed and adapted from bioinformatics. Processes such as pattern recognition, reasoning and problem solving are achieved via the building of multiple alignments, while unsupervised learning is achieved by creating patterns from sensory information and also by creating patterns from multiple alignments in which there is a partial match between one pattern and another. It is envisaged that, in SP-neural, short-lived neural structures equivalent to multiple alignments will be created via an inter-play of excitatory and inhibitory neural signals. It is also envisaged that unsupervised learning will be achieved by the creation of pattern assemblies from sensory information and from the neural equivalents of multiple alignments, much as in the non-neural SP theory—and significantly different from the “Hebbian” kinds of learning which are widely used in the kinds of artificial neural network that are popular in computer science. The paper discusses several associated issues, with relevant empirical evidence.

show abstract

“…Quian Quiroga et al [1] found that a large proportion of MTL neurons responded to both pictures and written names of particular individuals or objects and speculates that ''MTL neurons encode an abstract representation of the concept triggered by the stimulus.'' These abstract cells are often called concept cells and are further discussed in [4,5].…”

Section: The Evidence For Abstract Category and Concept Cells In The mentioning

confidence: 99%

On Findings of Category and Other Concept Cells in the Brain: Some Theoretical Perspectives on Mental Representation

Roy

2014

Cogn Comput

Self Cite

View full text Add to dashboard Cite

There is substantial neurophysiological evidence from decades of single-cell studies that category and other concept cells exist in both human and animal brains. This indicates that the brain can generalize and create abstract concepts and encode and represent such abstractions using single cells. These single-cell findings cannot be accounted for and explained by the theory of distributed representation and population coding, the dominant theory in the brain sciences. In light of these findings, this paper reexamines the two contending mental representational schemes, localist and distributed, on the basis of computational efficiency, the ability to simultaneously process and activate many different concepts, and the structure for semantic cognition. The evidence for category and concept cells favors localist representation in the brain.

show abstract

An extension of the localist representation theory: grandmother cells are also widely used in the brain

Cited by 13 publications

References 18 publications

Elements, structure and dynamics of the mind-brain system. A model based on 'Cognitive Feature Detectors'

Elements, structure and dynamics of the mind-brain system. A model based on 'Cognitive Feature Detectors'

Information Compression, Multiple Alignment, and the Representation and Processing of Knowledge in the Brain

On Findings of Category and Other Concept Cells in the Brain: Some Theoretical Perspectives on Mental Representation

Contact Info

Product

Resources

About