There are functional differences between the Neural Engineering Framework and standard artificial neural networks
There are notable functional and architectural differences between the Neural Engineering Framework (NEF) and standard artificial neural networks, specifically regarding biological plausibility, temporal coding, and semantic pointer architectures.
The retrieved papers consistently describe the Neural Engineering Framework (NEF) as a unique methodology for constructing spiking neural networks and cognitive architectures (like Spaun and robotic controllers) using principles like semantic pointers and biological constraints, which set it apart from standard artificial neural networks.
Andrew Mundy, James C. Knight, T. Stewart, S. Furber. An efficient SpiNNaker implementation of the Neural Engineering Framework. 2015. https://doi.org/10.1109/IJCNN.2015.7280390
Paper 0 details how the Neural Engineering Framework (NEF) builds functional neural systems with specific computational and memory challenges that differ from standard spiking network simulations.
See more details
Dailin Marrero, John Kern, Claudio Urrea. A Novel Robotic Controller Using Neural Engineering Framework-Based Spiking Neural Networks. 2024. https://doi.org/10.3390/s24020491
Paper 1 explains that the NEF utilizes temporal coding mechanisms and specific neuron models that provide functional advantages over conventional neural networks in robotic control tasks.
Jan Gosmann, C. Eliasmith. Optimizing Semantic Pointer Representations for Symbol-Like Processing in Spiking Neural Networks. 2016. https://doi.org/10.1371/journal.pone.0149928
Paper 2 demonstrates how the NEF realizes Semantic Pointer Architectures in spiking neural networks to handle symbol-like processing with distinct trade-offs.
Dumont NS, Furlong PM, Orchard J, Eliasmith C. Exploiting semantic information in a spiking neural SLAM system.. 2023. https://doi.org/10.3389/fnins.2023.1190515
Paper 4 presents a biologically plausible SLAM model built using NEF tools for large-scale cognitive modeling, employing vector representations and continuous attractors distinct from standard ANN architectures.
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