The authors draw together the results of a series of detailed computational studies and show how they are contributing to the development of a theory of hippocampal function. A new part of the theory introduced here is a quantitative analysis of how backprojections from the hippocampus to the neocortex could lead to the recall of recent memories. The theory is then compared with other theories of hippocampal function. First, what is computed by the hippocampus is considered. The hypothesis the authors advocate, on the basis of the effects of damage to the hippocampus and neuronal activity recorded in it, is that it is involved in the formation of new memories by acting as an intermediate-term buffer store for information about episodes, particularly for spatial, but probably also for some nonspatial, information. The authors analyze how the hippocampus could perform this function, by producing a computational theory of how it operates, based on neuroanatomical and neurophysiological information about the different neuronal systems con-tained within the hippocampus. Key hypotheses are that the CA3 pyramidal cells operate as a single autoassociation network to store new episodic information as it arrives via a number of specialized preprocessing stages from many association areas of the cerebral cortex, and that the dentate

1. To analyze the selectivity and the sparseness of firing to visual stimuli of single neurons in the primate temporal cortical visual area, neuronal responses were measured to a set of 68 visual stimuli in macaques performing a visual fixation task. The popula-tion of neurons analyzed had responses that occurred primarily to faces. The stimuli included 23 faces, and 45 nonface images of real-world scenes, so that the function of this brain region could be analyzed when it was processing natural scenes. 2. The neurons were selected to meet the previously used crite-ria of face selectivity by responding more than twice as much to the optimal face as to the optimal nonface stimulus in the set. Application of information theoretic analyses to the responses of these neurons confirmed that their responses contained much more information about which of 20 face stimuli had been seen (on average 0.4 bits) than about which (of 20) nonface stimuli had

The CA3 network in the hippocampus may operate as an autoassociator, in which declarative memories, known to be dependent on hippocampal processing, could be stored, and subsequently retrieved, using modifiable synaptic efficacies in the CA3 recurrent collateral system. On the basis of this hypothesis, the authors explore the computational relevance of the extrinsic afferents. to the CA3 network. A quantitative statistical analysis of the information that may be relayed by such afferent connections reveals the need for two distinct systems of input synapses. The synapses of the first system need to be strong (but not associatively modifiable) in order to force, during learning, the CA3 cells into a pattern of activity relatively independent of any inputs being received from the recurrent collaterals, and which thus reflects sizable amounts of new information. It is proposed that the mossy fiber system performs this function. A second system, with a large number of associatively modifiable synapses on each receiving cell, is needed in order to relay a signal specific enough to initiate the retrieval process. This may be identified, we propose, with the perforant path input to CA3. Key words: hippocampus, autoassociative memory, attractor neural networks, associative synapses, information storage

Abstract not found

The dynamics of a model attractor neural network, dominated by collateral feedback, composed of excitatory and inhibitory neurons described by afferent currents and spike rates, is studied analytically. The network stores stimuli learned in a temporal sequence. The statistical properties of the delay activities are investigated analytically under the approximation that no neuron is activated by more than one of the learned stimuli, and that inhibitory reaction is instantaneous. The analytic results reproduce the details of simulations of the model in which the stored memories are uncorrelated, and neurons can be shared, with low probability, by different stimuli. As such, the approximate analytic results account for delayed match to sample experiments of Miyashita in the inferotemporal cortex of monkeys. If the stimuli used in the experiment are uncorrelated, the analysis deduces the mean coding level f in a stimulus (i.e. the mean fraction of neurons activated by a given s...

Interpreting recent single unit recordings of delay activities in delayed match to sample experiments in anterior ventral temporal (AVT) cortex of monkeys in terms of reverberation dynamics, we present a model neural network of quasi-realistic elements which reproduces the empirical results in great detail. Information about the contiguity of successive stimuli in the training sequence, representing the fact that training is done on a set of uncorrelated stimuli presented in a fixed temporal sequence, in embedded in the synaptic structure. The model reproduces quite accurately the correlations between delay activity distributions corresponding to stimulation with the uncorrelated stimuli used for training. It reproduces also the activity distributions of spike rates on sample cells as a function of the stimulating pattern. It is, in our view, the first time that a computational phenomenon, represented on the neuro-physiological level is reproduced in all its quantitative aspects. The m...

this paper with describing linked temporal sequences of events.) The hippocampus is also implicated in spatial memory. For example, damage to the hippocampal system in monkeys produces de# cits in learning about where objects are and where responses must be made (Rolls 1996; Gaffan 1998), and in rats produces spatial learning de# cits (Martin et al. 2000). Neurophysiologically, hippocampal neurons in rats respond to the place where the animal is located (O'Keefe 1990; Kubie & Muller 1991; Wilson & McNaughton 1993), and in primates to the place being viewed (Rolls et al. 1997; Rolls 1999). It has thus been a long-standing question about whether the hippocampus and nearby temporal lobe structures are involved in episodic memory or spatial function. In this paper we show that this question can be resolved by revealing that a single neural network can implement both episodic and spatial memory

I discuss the construction of models that describe the firing rates of excitatory and inhibitory neurons in biological neural networks. A model is presented that incorporates both slow linear and fast nonlinear inhibition. With the appropriate excitatory-to-excitatory couplings this model can act as an associative memory in which pattern recognition is signalled by resonant firing behavior. Stored memories are represented by fixed points of the excitatory and fast inhibitory dynamics. After memory recovery, slow inhibition returns the system to the silent, resting state. Published in Benhar, O., Bosio, C., Del Giudice, P. and Tabet, E., eds. Neural Networks: From Biology to High- Energy Physics (ETS Editrice, Pisa, 1991) pp. 179-196. 1. INTRODUCTION The model neural networks used for pattern recognition and data analysis are inspired by, but only impressionistically related to, their biological counterparts. A central issue in the biological study of neural networks is whether the in...

We investigate the properties of an autoassociative network of thresholdlinear units whose synaptic connectivity is spatially structured and asymmetric. Since the methods of equilibrium statistical mechanics cannot be applied to such a network due to the lack of a Hamiltonian, we approach the problem through a signal-to-noise analysis, that we adapt to spatially organized networks. The conditions are analyzed for the appearance of stable, spatially non-uniform profiles of activity with large overlaps with one of the stored patterns. It is also shown, with simulations and analytic results, that the storage capacity does not decrease much when the connectivity of the network becomes short range. In addition, the method used here enables us to calculate exactly the storage capacity of a randomly connected network with arbitrary degree of dilution. 1

Interpreting recent single unit recordings of delay activities in delayed match to sample experiments in AVT cortex of monkeys in terms of reverberation dynamics, we present a model neural network of quasi-realistic elements which reproduces the empirical results in great detail. Including, in the synaptic structure information about the contiguity of successive stimuli in the training sequence, to represent the fact that training is done on a set of uncorrelated stimuli presented in a fixed temporal sequence, the model reproduces quite accurately the correlations between delay activity distributions corresponding to stimulation with the uncorrelated stimuli used for training. It reproduces also the activity distributions of spike rates on sample cells as a function of the stimulating pattern. It is, in our view, the first time that a computational phenomenon, represented on the neuro-physiological level is reproduced in all its quantitative aspects. The model is then used to make pred...