Neural Network Architecture of Embodied Intelligence

Main Article Content

Abstract

In recent years, advances in artificial intelligence (AI) and machine learning have been driven by advances in the development of large language models (LLMs) based on deep neural networks. At the same time, despite its substantial capabilities, LLMs have fundamental limitations such as spontaneous unreliability in facts and judgments; making simple errors that are dissonant with high competence in general; credulity, manifested by a willingness to accept a user's knowingly false claims as true; and lack of knowledge about events that have occurred after training has been completed.


Probably the key reason is that bioinspired intelligence learning occurs through the assimilation of implicit knowledge by an embodied form of intelligence to solve interactive real-world physical problems. Bioinspired studies of the nervous systems of organisms suggest that the cerebellum, which coordinates movement and maintains balance, is a prime candidate for uncovering methods for realizing embodied physical intelligence. Its simple repetitive structure and ability to control complex movements offer hope for the possibility of creating an analog to adaptive neural networks.


This paper explores the bioinspired architecture of the cerebellum as a form of analog computational networks capable of modeling complex real-world physical systems. As a simple example, a realization of embodied AI in the form of a multi-component model of an octopus tentacle is presented, demonstrating the potential in creating adaptive physical systems that learn and interact with the environment.

Article Details

References

1. Bubeck S. et al. Sparks of artificial general intelligence: Early experiments with gpt-4 // arXiv preprint arXiv:2303.12712. 2023.
2. Achiam J. et al. Gpt-4 technical report // arXiv preprint arXiv:2303.08774. 2023.
3. Brown T.B. Language models are few-shot learners // arXiv preprint arXiv:2005.14165. 2020.
4. TOP 500. June 2024 URL: https://www.top500.org/lists/top500/2024/06/. 2024
5. Paul A. OpenAI argues it is ‘impossible’ to train ChatGPT without copyrighted work // Popular Science. 2024.
6. Olivera F. Memory systems in organizations: an empirical investigation of mechanisms for knowledge collection, storage and access // Journal of Management Studies. 2000. Vol. 37, No. 6. P. 811–832.
7. Chen H., Vondrick C., Mao C. Selfie: Self-interpretation of large language model embeddings // arXiv preprint arXiv:2403.10949. 2024.
8. Marks S., Tegmark M. The geometry of truth: Emergent linear structure in large language model representations of true/false datasets // arXiv preprint arXiv:2310.06824. 2023.
9. Abramski K. et al. Cognitive network science reveals bias in gpt-3, gpt-3.5 turbo, and gpt-4 mirroring math anxiety in high-school students // Big Data and Cognitive Computing. 2023. Vol. 7, No. 3. P. 124.
10. Crowder J.A. et al. Implicit learning in artificial intelligence // Artificial Psychology: Psychological Modeling and Testing of AI Systems. 2020. P. 139–147.
11. Seger C.A. Implicit learning // Psychological bulletin. 1994. Vol. 115, No. 2. P. 163–196.
12. Olivera F. Memory systems in organizations: an empirical investigation of mechanisms for knowledge collection, storage and access // Journal of Management Studies. 2000. Vol. 37, No. 6. P. 811–832.
13. Rajan S., Pantelyat A. Apraxias // eLS. 2018. P. 1–11.
14. Goldenberg G. Apraxia and the parietal lobes // Neuropsychologia. 2009. Vol. 47, No. 6. P. 1449–1459.
15. Gardner H.E. Frames of mind: The theory of multiple intelligences. Basic books, 2011.
16. Sitti M. Physical intelligence as a new paradigm // Extreme Mechanics Letters. 2021. Vol. 46. P. 101340.
17. Федеральная служба государственной статистики (2021) Рабочая сила, занятость и безработица в России в 2021. Статистический сборник. Таблица 2.33.
18. Nvidia CEO Jensen Huang says robots are the next wave of AI 2024
19. Duan J. et al. A survey of embodied ai: From simulators to research tasks // IEEE Transactions on Emerging Topics in Computational Intelligence. 2022. Vol. 6, No. 2. P. 230–244.
20. Wilson R.A., Foglia L. Embodied Cognition, in «The Stanford Encyclopedia of Philosophy». 2017.
21. Cangelosi A. et al. Embodied intelligence // Springer handbook of computational intelligence. 2015. P. 697–714.
22. Jékely G., Keijzer F., Godfrey-Smith P. An option space for early neural evolution // Philosophical Transactions of the Royal Society B: Biological Sciences. 2015. Vol. 370, No. 1684. P. 20150181.
23. Иргер И.М. Клиника и хирургическое лечение опухолей мозжечка. М.: Медгиз, 1959.
24. Itō M. The cerebellum and neural control. Raven press, 1984.
25. Kawato M. et al. 50 years since the Marr, Ito, and Albus models of the cerebellum // Neuroscience. 2021. Vol. 462. P. 151–174.
26. Kalinichenko S.G., Motavkin P.A. Cerebellar Cortex; Institute Marine Biology FEB RAS. Moscow: Nauka, 2005.
27. Tanaka H. et al. The cerebro-cerebellum as a locus of forward model: a review // Frontiers in systems neuroscience. 2020. Vol. 14. P. 19.
28. Miall R.C., Wolpert D.M. Forward models for physiological motor control // Neural networks. 1996. Vol. 9, No. 8. P. 1265–1279.
29. Shadmehr R., Smith M.A., Krakauer J.W. Error correction, sensory prediction, and adaptation in motor control // Annual review of neuroscience. 2010. Vol. 33, No. 1. P. 89–108.
30. Habas C. et al. Deep cerebellar nuclei (DCN) and language // The linguistic cerebellum. Academic Press, 2016. P. 337–353.
31. Kalman R.E. A New Approach to Linear Filtering and Prediction Problems // Journal of Basic Engineering. 1960. Vol. 82, No. 1. P. 35.
32. Thomas A. Cerebellar functions [Electronic resource] / André Thomas. New York: Journal of Nervous and Mental Disease Pub. Co. 1912. P. 244, Figure 67 (Monograph series; issue 12).
33. Herculano-Houzel S. Coordinated scaling of cortical and cerebellar numbers of neurons // Frontiers in neuroanatomy. 2010. Vol. 4. P. 952.
34. Funahashi K., Nakamura Y. Approximation of dynamical systems by continuous time recurrent neural networks // Neural networks. 1993. Vol. 6, No. 6. P. 801–806.
35. Арнольд В.И. Обыкновенные дифференциальные уравнения. МЦНМО. 2014. C. 5.
36. Consalez G.G. et al. Origins, development, and compartmentation of the granule cells of the cerebellum // Frontiers in neural circuits. 2021. Vol. 14. P. 611841.
37. Universidad Complutense de Madrid. Cuadros originales delineados por Santiago Ramón y Cajal y pintados por Ramón Padró. Corteza cerebelosa. URL: https://www.ucm.es/seccionbiocel/historia
38. De Schutter E., Vos B., Maex R. The function of cerebellar Golgi cells revisited // Progress in brain research. 2000. Vol. 124. P. 81–93.
39. D'Angelo E. et al. The cerebellar Golgi cell and spatiotemporal organization of granular layer activity // Frontiers in neural circuits. 2013. Vol. 7. P. 93.
40. Maass W. Networks of spiking neurons: the third generation of neural network models // Neural networks. 1997. Vol. 10, No. 9. P. 1659–1671.
41. Izhikevich E.M., Hoppensteadt F. Classification of bursting mappings // International Journal of Bifurcation and Chaos. 2004. Vol. 14, No. 11. P. 3847–3854.
42. Izhikevich E.M. Dynamical systems in neuroscience: The geometry of excitability and bursting. MIT press, 2007.
43. Small J.S. The analogue alternative: The electronic analogue computer in Britain and the USA, 1930–1975. Routledge, 2013.
44. Чикуров Н.Г. Численный метод решения обыкновенных дифференциальных уравнений с помощью приведения их к форме Шеннона // Матем. моделирование. 2020. Т. 32, № 8. С. 3–20.
45. Marullo C., Agliari E. Boltzmann machines as generalized Hopfield networks: a review of recent results and outlooks // Entropy. 2020. Vol. 23, No. 1. P. 34.
46. Nurutdinov A.R., Latypov R.K. A Novel Artificial Neural Networks Architecture Based on Cerebellum Model // Lobachevskii Journal of Mathematics. 2023. Vol. 44, No. 2. P. 733–746.
47. Nurutdinov A., Latypov R. Dynamically Predicting Wi-Fi Coverage Mapping Using Bioinspired Neural Networks // 2023 International Conference on Software, Telecommunications and Computer Networks (SoftCOM). IEEE, 2023. P. 1–6.
48. Gupta A. et al. Embodied intelligence via learning and evolution // Nature communications. 2021. Vol. 12, No. 1. P. 5721.
49. Zullo L. et al. Motor control pathways in the nervous system of Octopus vulgaris arm // Journal of Comparative Physiology A. 2019. Vol. 205. P. 271–279.
50. Ландау Л.Д., Лифшиц Е.М. Теоретическая физика. Издание 5-е, стереотипное. М.: Физматлит. 2004. Т. I. Механика. С. 26.
51. Cerebellum cytoarchitecture for solving the equation of harmonic oscillations. URL: https://github.com/airatn/Cerebellum-cytoarchitecture-for-solving-the- equation-of-harmonic-oscillations
52. Муханов К.К. Область применения и классификации ферм // Металлические конструкции. 1976. C. 287–293.
53. Перельмутер А.В. Очерки по истории металлических конструкций. Издательство АСВ, 2012.
54. ferms. URL: https://youtu.be/GVK1D8-qolE
55. Chang H.S. et al. Energy-shaping control of a muscular octopus arm moving in three dimensions // Proceedings of the Royal Society A. 2023. Vol. 479, No. 2270. P. 20220593.
56. tentacle. URL: https://youtu.be/en6EwpKJr6U
57. Wilson R.A., Foglia L. Embodied Cognition, in «The Stanford Encyclopedia of Philosophy», 2017.
58. Godfrey-Smith P. Other minds: The octopus, the sea, and the deep origins of consciousness. Farrar, Straus and Giroux, 2016.