Artificial neural network weights directly correspond to biological synaptic strengths.
Artificial neural networks were originally inspired by the brain, but their mathematical weights do not directly correspond to biological synaptic strengths.
The retrieved papers demonstrate that while artificial neural networks (ANNs) draw mathematical inspiration from biological nervous systems or serve as computational models and tools to analyze brain data, they are distinct engineering constructs. ANNs utilize abstract numerical weights optimized via algorithms like backpropagation, whereas biological synapses undergo complex biochemical, molecular, and electrophysiological processes (such as nonlinear inhibitory plasticity and spike-timing-dependent plasticity) that do not directly map to standard artificial weights.
Christian L. Ebbesen, Robert C. Froemke. Automatic mapping of multiplexed social receptive fields by deep learning and GPU-accelerated 3D videography. 2020. https://doi.org/10.1101/2020.05.21.109629
Paper 0 uses deep learning as a tool to analyze recorded neural data rather than equating artificial weights to biological synapses.
See more details
Christoph Miehl, Julijana Gjorgjieva. Stability and learning in excitatory synapses by nonlinear inhibitory plasticity. 2022. https://doi.org/10.1101/2022.03.28.486052
Paper 1 discusses biological synaptic plasticity and excitatory/inhibitory balance as distinct physiological mechanisms.
Aishwarya H. Balwani, Eva L. Dyer. A Deep Feature Learning Approach for Mapping the Brain’s Microarchitecture and Organization. 2020. https://doi.org/10.1101/2020.05.26.117473
Paper 2 uses convolutional neural networks to model and map anatomical brain structures rather than claiming artificial weights mirror biological synapses.
David Abramian, Anders Eklund, Evren Özarslan. Super-resolution mapping of anisotropic tissue structure with diffusion MRI and deep learning. 2023. https://doi.org/10.1101/2023.04.04.535586
Paper 3 applies deep learning to improve the resolution of diffusion MRI data for mapping axonal fiber tracts rather than equating network parameters to biological strengths.
Qais Yousef, Pu Li. Synaptic plasticity-based regularizer for artificial neural networks. 2024. https://doi.org/10.21203/rs.3.rs-4114689/v1
Paper 4 draws inspiration from neuroplasticity to design regularizers for artificial neural networks, treating biological concepts as algorithmic inspiration rather than direct equivalence.
V. A. Kulagin. REINFORCEMENT LEARNING OF SPIKING NEURAL NETWORKS USING TRACE VARIABLES FOR SYNAPTIC WEIGHTS WITH MEMRISTIVE PLASTICITY. 2025. https://doi.org/10.7868/s3034548025030033
Paper 5 explores spiking neural networks implemented with memristors as hardware approximations, but treats them as engineered models rather than direct biological equivalents.
Najib J. Majaj, Denis G. Pelli. Deep learning: Using machine learning to study biological vision. 2017. https://doi.org/10.1101/178152
Paper 6 notes that while artificial neural networks were originally inspired by the brain and serve as models of brain function, they operate as statistical and computational systems distinct from direct biological correspondence.
The paper trail · every fact has a biography
Challenge the receipt
Citation formatting by citeproc-js (Frank Bennett) and the Citation Style Language project. Source and licenses.
Terms · Privacy · How verdicts work · Dispute this receipt