Simon Schultz, Alessandro Treves
Published 1997-09-19, updated 1997-11-09Version 2
The information that a pattern of firing in the output layer of a feedforward network of threshold-linear neurons conveys about the network's inputs is considered. A replica-symmetric solution is found to be stable for all but small amounts of noise. The region of instability depends on the contribution of the threshold and the sparseness: for distributed pattern distributions, the unstable region extends to higher noise variances than for very sparse distributions, for which it is almost nonexistant.
Comments: 19 pages, LaTeX, 5 figures. Also available at http://www.mrc-bbc.ox.ac.uk/~schultz/papers.html . Submitted to Phys. Rev. E Minor changes
Phase Transitions of Neural Networks
Retrieval dynamics of neural networks for sparsely coded sequential patterns
Stochastic learning in a neural network with adapting synapses
