Abstract
The computational resources of a neuromorphic network model introduced earlier are investigated in the context of such hierarchical systems as the mammalian visual cortex. It is argued that a form of ubiquitous spontaneous local convolution, driven by spontaneously arising wave-like activity—which itself promotes local Hebbian modulation—enables logical gate-like neural motifs to form into hierarchical feed-forward structures of the Hubel-Wiesel type. Extra-synaptic effects are shown to play a significant rôle in these processes. The type of logic that emerges is not Boolean, confirming and extending earlier findings on the logic of schizophrenia.
Similar content being viewed by others
Data availability
No datasets were generated or analysed during the current study.
References
Aggarwal, C. S. (2018). Neural networks and deep learning: A textbook. Berlin, Heidelberg, New York: Springer-Verlag.
Arthur, W. (2023). D’Arcy Thompson’s morphological transformations: Issues of causality and dimensionality. Biological Theory. https://doi.org/10.1007/s13752-023-00438-0
Dempsey, W. P., Du, Z., Natcochiy, A., Smith, D. K., Czajkowski, A. A., Robson, D. N., Li, J. M., Applebaum, S., Truong, T. V., Kesselman, C., Fraser, S. E., & Arnold, D. B. (2022). Regional synapse gain and loss accompany memory formation in larval Zebrafish. PNAS, 3, e2107661119. https://doi.org/10.1073/pnas.2107661119
Enock, M., & Schwartz, J.-M. (1992). Kac algebras and the duality of locally compact groups. Berlin, Heidelberg, New York: Springer-Verlag.
Fernandez-Ruiz, A., Sirota, A., Lopes-dos-Santos, V., & Dupret, D. (2023). Over and above frequency: Gamma oscillations as units of neural circuit operations. Neuron. https://doi.org/10.1016/j.neuron.2023.02.026
Fries, P. (2009). Neuronal gamma-band synchronization as a fundamental process in cortical computation. Annual Review of Neuroscience, 32, 209–224. https://doi.org/10.1146/annurev.neuro.051508.135603
Gilbert, S. F. (2007). Fate maps, gene expression maps, and the evidentiary structure of evolutionary developmental biology. In D. M. Laublicher & J. Maienschein (Eds.), From embryology to evo-devo: A history of developmental evolution. Cambridge: MIT Press.
Hewitt, E., & Ross, K. A. (1963). Abstract harmonic analysis I. Berlin, Heidelberg, New York: Springer-Verlag.
Hewitt, E., & Ross, K. A. (1970). Abstract harmonic analysis II. Berlin, Heidelberg, New York: Springer-Verlag.
Hubel, D., & Wiesel, T. (1962). Receptive fields, binocular interaction and functional architecture in the cat’s visual cortex. Journal of Physiology, 160, 106–154.
Johansen, J. P., Diaz-Mataix, L., Hamanaka, H., Ozawa, T., Ycu, E., Koivumaa, J., Kumar, A., Hou, M., Deisseroth, K., Boyden, E. S., & LeDoux J. E. (2014). Hebbian and neuromodulatory mechanisms interact to trigger associative memory formation. Proceedings of the National Academy of Sciences of the USA, 111(51), E5584–ESS92. https://doi.org/10.1073/pnas.1421304111. Epub 2014 Dec 8. PMID: 25489081; PMCID: PMC4280619.
LeCun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998). Gradient-based learning applied to document recognition. Proceedings of the IEEE, 86(11), 2278–2324.
Lockhofen, D. E. L., & Mulert, C. (2021). Neurochemistry of visual attention. Frontiers in Neuroscience, 15, 643597. https://doi.org/10.3389/fnins.2021.643597
Pinotsis, D. A., Fridman, G., & Miller, E. K. (2023). Cytoelectric coupling: Electric fields sculpt neural activity and tune the brain’s infrastructure. Progress in Neurobiology, 226, 102465. https://doi.org/10.1016/j.pneurobio.2023.102465. ISSN 0301–0082.
Poggio, T. A., & Anselmi, F. (2016). Visual cortex and deep networks. Cambridge, MA, and London: The MIT Press.
Schaffner, J., Dongqi Bao, S., Tobler, P. N., Hare, T. A., & Polanian, R. (2023). Sensory perception relies on fitness-maximizing codes. Nature Human Behavior. https://doi.org/10.1038/s41562-023-01584-y
Selesnick, S. A. (2019). Tsien’s power-of-two law in a neuromorphic network model suitable for artificial intelligence. IfCoLog Journal of Logics and their Applications, 6(7), 1223–1251.
Selesnick, S. A. (2022). Quantum-like networks. An approach to neural behavior through their mathematics and logic: World Scientific.
Selesnick, S. (2023). Neural waves and short term memory in a neural network model. Journal of Biological Physics, 49, 159–194. https://doi.org/10.1007/s10867-023-09627-1
Selesnick, S. A., & Owen, G. S. (2012). Quantum-like logics and schizophrenia. Journal of Applied Logic, 10(1), 115–126. https://doi.org/10.1016/j.jal.2011.12.001
Selesnick, S. A., & Piccinini, G. (2018). Quantum-like behavior without quantum physics II. A quantum-like model of neural network dynamics. Journal of Biological Physics, 44, 501–538. https://doi.org/10.1007/s10867-018-9504-9
Selesnick, S. A., & Piccinini, G. (2019). Quantum-like behavior without quantum physics III. Logic and memory. Journal of Biological Physics, 45, 335–366. https://doi.org/10.1007/s10867-019-09532-6
Selesnick, S. A., Rawling, J. P., & Piccinini, G. (2017). Quantum-like behavior without quantum physics I. Kinematics of neural-like systems. Journal of Biological Physics, 43, 415–444. https://doi.org/10.1007/s10867-017-9460-9
Silverstein, S. M., & Rosen, R. (2015). Schizophrenia and the eye. Schizophrenia Research: Cognition, 2(2), 46–55. https://doi.org/10.1016/j.scog.2015.03.004
Acknowledgements
Many thanks are owed to Piers Rawling for probabilistic help among many other useful discussions, to Ronald Munson for invaluable advice, and to Wallace Arthur for his insights into the two dimensions of biological time, among other things.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Author information
Authors and Affiliations
Contributions
Stephen Selesnick is responsible for 100% of this article.
Corresponding author
Ethics declarations
Informed consent
The author declares that there were no human or animal participants involved in this purely theoretical study.
Conflict of interest
The author declare no conflict of interest.
Additional information
Action Editor: Pulin Gong
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendices
1.1 A. The multiplicative rules of the Gentzen sequent calculus GN
GN
Structural Rules
Exchange
Weakening
Contraction
The Identity Group
Axiom
Cut
Multiplicative Logical Rules
Conjunctive (Multiplicative) Connective
Of Course operator
!
Capital Greeks stand for finite sequences of formulas including possibly the empty one, and D stands for either a single formula or no formula, i.e. the empty sequence, and when it appears in the form \(\otimes D\), the \(\otimes\) symbol is presumed to be absent when D is empty. If \(\Gamma\) denotes the sequence \(A_1,A_2, \ldots , A_n\) then \(!\Gamma\) will denote the sequence \(!A_1,!A_2, \ldots , !A_n\). The Girard of course exponential operator ! is sometimes pronounced “bang.” The sign \(\otimes\) should strictly speaking just be regarded as an abstract symbol though there is no harm and large benefit if the reader just thinks of this and all the other symbols involved as pertaining directly to the category of finite dimensional real Hilbert spaces. If this is done, then commas are replaced by \(\otimes\), \(A \vdash B\) is replaced by a linear map \(A \rightarrow B\), !A is replaced by the exterior algebra over the Hilbert space A, and blank spaces by \(\mathbb {R}\). In applications, one adds “non-logical” axioms to the GN rules, for instance to depict a brain as a family of linked networks.
1.2 B. A note on convolution
An associative algebra B over the ring R may be defined as follows. B is a module over R with a map \(m: B \otimes B \rightarrow B\) called multiplication (or sometimes abusively product). With \(m(a\otimes b):= ab\) associativity means \(a(bc) = (ab)c\). This condition may be expressed in the form of a commutative diagram specifying that \(m(1_B \otimes m) = m (m\otimes 1_B)\). One virtue of expressing this condition diagramatically is that it may easily be dualized by reversing the arrows.
A coassociative colagebra A over the ring R is a module over R with a map \(\psi : A \rightarrow A \otimes A\) called comultiplication (or sometimes abusively coproduct) with a dual version of the diagram above commuting, namely \((\psi \otimes 1_A) \psi = (1_A \otimes \psi ) \psi\).
Let A be such a coalgebra and B such an algebra. Given two R-module maps \(L, M: A \rightarrow B\) we get another module map \(L*M:A \rightarrow B\) as follows.
This *-operation is called convolution. It is easily seen to be an associative product, due to the associativity of B and the coassociativity of A.
The simplest way to see that ordinary convolution may be expressed in this form is to consider the group algebra R[G] of a finite group G written multiplicatively. This is just the space of functions on G into R, and here we ignore the algebra product on it. Each such function may be expressed in the form
where \(a_g \in R\) and \(\chi _g\) is the characteristic function on G which is 1 on g and zero elsewhere. We have a coassociative comultiplication on R[G] induced from the associative group multiplication \(G \times G \rightarrow G\), namely the map \(\psi : R[G] \rightarrow R[G \times G] \cong R[G] \otimes R[G]\) given for \(f \in R[G]\) by \(\psi (f)(g_1,g_2) = f(g_1g_2)\).
We can express this precisely in tensorial form for \(f = \chi _{g_0}\) as follows.
Now take B to be R and \(m: R\otimes R\rightarrow R\) to be multiplication in R and note that any module map \(L: R[G] \rightarrow R\) acting on some \(f = \sum _{g \in G} a_g \chi _g\) takes the form \(L(f) = \sum _{g \in G} a_g L(\chi _g)\). Of course, any such L may be regarded as a function on G via the one-to-one map \(g \leftrightarrow \chi _g\) so that we may write \(L(g):= L(\chi _g)\) and concomitantly \(L(f) = \sum _g a_g L(g)\). Then we have
so that the function of \(h\in G\) corresponding to the functional \(L*M\) is
This is just the usual convolution of the functions L and M. This construction has generalizations to measures and functions on locally compact groups and in other areas of harmonic analysis.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Selesnick, S. Neural waves and computation in a neural net model I: Convolutional hierarchies. J Comput Neurosci 52, 39–71 (2024). https://doi.org/10.1007/s10827-024-00866-2
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10827-024-00866-2