1,721,120 research outputs found
Graded-response neurons and information encodings in autoassociative memories
No full textA general mean-field theory is presented for an attractor neural network in which each elementary unit is described by one input and one output real variable, and whose synaptic strengths are determined by a covariance imprinting rule. In the case of threshold-linear units, a single equation is shown to yield the storage capacity for the retrieval of random activity patterns drawn from any given probability distribution. If this distribution produces binary patterns, the storage capacity is essentially the same as for networks of binary units. To explore the effects of storing more structured patterns, the case of a ternary distribution is studied. It is shown that the number of patterns that can be stored can be much higher than in the binary case, whereas the total amount of retrievable information does not exceed the limit obtained with binary patterns. © 1990 The American Physical Society
Mean-Field analysis of neuronal spike dynamics
No full textI consider a mean-field description of the dynamics of interacting intergrate-and-fire neuron-like units. The basic dynamical variables are the membrane potential of each (point-like) 'cell' and the conductance associated with each synaptic connection, both of which evolve discontinuously in time. In addition, an intrinsic potassium conductance, also evolving discontinuously in time, can be associated to each cell in order to model firing frequency adaptation in real neurons. The mean-field theory is exact if the units can be grouped into NC classes, each comprising infinitely many identical, and identically coupled, units; and can be used as an approximation if, instead, a class comprises few or just one unit. The formalism yields both the stationary asynchronous solutions and the transients leading to those solutions. The full spectrum of time-constants for the transients associated with one particular steady state is given by a single equation, imposing the vanishing of the determinant of an NC×NC matrix. In the case of an associative memory, this equation can be manipulated into a simple form, using standard replica methods. An analysis of the spectrum indicates that the major role in determining the transients time constants is played by the effective decay times of postsynaptic currents, which can be quite short. This suggests that local recurrent neocortical circuits may produce a very rapid dynamics, consistent with such circuits participating in the rapid course of information processing, evidenced by new experimental data recorded in primate temporal cortex. © 1993 Informa UK Ltd All rights reserved: reproduction in whole or part not permitted
Threshold-linear formal neurons in auto-associative nets
No full textMost analytical results concerning the long-time behaviour of associative memory networks have been obtained by using binary elementary units. The use of alternative types of neuron-like processing elements is considered as a way of testing the generality of those results and of approaching biological realism. In particular, threshold-linear units are proposed as appropriate in models designed to reproduce low firing rates, in which long-time stability does not rely on single unit saturation. Such units are simple enough to allow detailed analytical understanding of the properties of the network. This is demonstrated by analysing the attractor states of a network operating at low rates. It is shown that while the interesting retrieval behaviour persists, the roles of the different parameters as well as the nature of the stable states change completely with respect to the binary implementation
Are spin-glass effects relevant to understanding realistic auto-associative networks?
No full textElementary units characterized by a threshold-linear (graded) response have been argued to model single neurons in auto-associative networks more realistically than binary units. The different way local activity is constrained in the two representations is shown here to have important consequences for the spin-glass-like properties of otherwise equivalent systems. In particular, in contrast with their binary counterparts, the threshold-linear Sherrington-Kirkpatrick model is stable with respect to replica symmetry-breaking (RSB), while threshold-linear fully connected neural networks with covariance learning are RSB unstable only in a very restricted region of their phase diagram. Whether or not spin-glass effects dominate attractor dynamics is suggested to affect considerably, among other things, the ability of auto-associative memories to encode new information
Dilution and sparse coding in threshold-linear nets
No full textThe storage capacity of an autoassociative memory with extremely diluted connectivity and with threshold-linear elementary units is studied in its dependence on the graded structure and on the sparseness of the coding scheme, and on the form of the learning rule used. As the coding becomes sparse, more patterns can be stored, and the difference in capacity (measured for a given number of modifiable synapses per unit) between fully connected and highly diluted systems vanishes. Graded (non-binary) codings, especially when used with learning rules nonlinear in their post-synaptic factor, further increase the number of patterns that can be stored by making their retrieved representation even sparser
More cortex, yes, but what flavour?
No full textcommentary to the target article by Francisco Aboitiz, Daniver Morales and Juan
Montiel, The evolutionary Origin of the Mammalian Isocortex: towards an Integrated
Developmental and Functional Approac
Mere functional characterization is not enough to understand memory circuits
No full textWhat exactly is going on via fornical connections? Aggleton and Brown's target article correctly stresses their importance, but a detailed understanding of their role in memory appears to require fresh research approaches
- …
