Free association transitions in models of cortical latching dynamics

Potts networks, in certain conditions, hop spontaneously from one discrete attractor state to another, a process we have called latching dynamics. When continuing indefinitely, latching can serve as a model of infinite recursion, which is nontrivial if the matrix of transition probabilities presents a structure, i.e. a rudimentary grammar. We show here, with computer simulations, that latching transitions cluster in a number of distinct classes: effectively random transitions between weakly correlated attractors; structured, history-dependent transitions between attractors with intermediate correlations; and oscillations between pairs of closely overlapping attractors. Each type can be described by a reduced set of equations of motion, which, once numerically integrated, matches simulations results. We propose that the analysis of such equations may offer clues on how to embed meaningful grammatical structures into more realistic models of specific recursive processes.

[1]  Noam Chomsky,et al.  The faculty of language: what is it, who has it, and how did it evolve? , 2002, Science.

[2]  Alessandro Treves,et al.  Uninformative memories will prevail: The storage of correlated representations and its consequences , 2007, HFSP journal.

[3]  Friedemann Pulvermüller,et al.  A neuronal model of the language cortex , 2007, Neurocomputing.

[4]  Harold L. Dibble,et al.  On the Emergence of Modern Humans , 1990, Current Anthropology.

[5]  Alessandro Treves,et al.  The storage capacity of Potts models for semantic memory retrieval , 2005 .

[6]  Lawrence W. Barsalou,et al.  Continuity of the conceptual system across species , 2005, Trends in Cognitive Sciences.

[7]  Alessandro Treves,et al.  Frontal latching networks: a possible neural basis for infinite recursion , 2005, Cognitive neuropsychology.

[8]  A. Braun,et al.  Toward an evolutionary perspective on conceptual representation: species-specific calls activate visual and affective processing systems in the macaque. , 2004, Proceedings of the National Academy of Sciences of the United States of America.

[9]  Prof. Dr. Valentino Braitenberg,et al.  Anatomy of the Cortex , 1991, Studies of Brain Function.

[10]  Kanter.  Potts-glass models of neural networks. , 1988, Physical review. A, General physics.

[11]  P. Manger,et al.  Order‐specific quantitative patterns of cortical gyrification , 2007, The European journal of neuroscience.

[12]  M. Tsodyks,et al.  The Enhanced Storage Capacity in Neural Networks with Low Activity Level , 1988 .

[13]  J. Fuster.  Memory in the cerebral cortex : an empirical approach to neural networks in the human and nonhuman primate , 1996 .

[14]  Marius Usher,et al.  A neural model of the dynamic activation of memory , 1993, Biological Cybernetics.

[15]  Ronald Cools,et al.  Mean-field theory for the Q-state Potts-glass neural network with biased patterns , 1993 .

[16]  R. Barton.  Evolutionary specialization in mammalian cortical structure , 2007, Journal of evolutionary biology.

[17]  Alessandro Treves,et al.  The complexity of latching transitions in large scale cortical networks , 2006, Natural Computing.