arXiv · 2103.14324
Eigenvalue spectrum of neural networks with arbitrary Hebbian length
Abstract
Associative memory is a fundamental function in the brain. Here, we generalize the standard associative memory model to include long-range Hebbian interactions at the learning stage, corresponding to a large synaptic integration window. In our model, the Hebbian length can be arbitrarily large. The spectral density of the coupling matrix is derived using the replica method, which is also shown to be consistent with the results obtained by applying the free probability method. The maximal eigenvalue is then obtained by an iterative equation, related to the paramagnetic to spin glass transition in the model. Altogether, this work establishes the connection between the associative memory with arbitrary Hebbian length and the asymptotic eigen-spectrum of the neural-coupling matrix.
Explore related subjects
Keep this discovery
Jianwen Zhou, Zijian Jiang, Tianqi Hou, Ziming Chen, K Y Michael Wong, Haiping Huang. 2021-03-26. Eigenvalue spectrum of neural networks with arbitrary Hebbian length. https://doi.org/10.1103/physreve.104.064307
Cite the original work for its findings. Save a collection to share your selection of sources.