Showing papers by "Edmund T. Rolls published in 1991"

What determines the capacity of autoassociative memories in the brain

Abstract: Threshold-linear (graded response) units approximate the real firing behaviour of pyramidal neurons in a simplified form, suited to the analytical study of large autoassociative networks. Here we extend previous results on threshold-linear networks to a much larger class of models, by considering different connectivities (including full feedback, highly diluted and multilayer feedforward architectures), different forms of Hebbian learning rules, and different distributions of firing rates (including realistic, continuous distributions of rates). This allows an evaluation of the main factors which may affect, in real cortical networks, the capacity for storage and retrieval of discrete firing patterns.In each case a single equation is derived, which determines both αc, the maximum number of retrievable patterns per synapse, and Im, the maximum amount of retrievable information per synapse. It is shown that: 1. Non-speeific effects, such as those usually ascribed to inhibition, or to neuramodulatary afferen...

Allocentric and egocentric spatial information-processing in the hippocampal-formation of the behaving primate

Abstract: To investigate how space is represented in the primate hippocampus, the activity of single neurons was recorded in the hippocampus of behaving macaque monkeys. Neurons that responded differently according to the position on a screen in which a stimulus was shown when the monkey had to remember the stimulus and its position were analyzed for their spatial fields. By moving the monkey into different positions relative to the screen, and the screen to different positions in the room, it was possible to separate neuronal representations in egocentric coordinates (i.e., defined relative to the monkey’s body axis) from representations in local allocentric coordinates (i.e., defined relative to the frame of reference provided by the screen on which the stimuli were displayed) and from representations in absolute allocentric coordinates (i.e., defined relative to absolute position in the room). It was found that 44% of the spatial neurons analyzed in these experiments responded in relation to space defined by the local frame of reference and not in relation to space defined in egocentric or in absolute allocentric coordinates. Two percent of the neurons responded in relation to the absolute position of a stimulus in the room (in absolute allocentric coordinates). Ten percent responded in relation to egocentric coordinates. Forty-four percent responded to a combination of the different coordinate systems investigated, including 23% that were shown to include an absolute allocentric component. Thus the main finding was that in the primate hippocampus many spatial cells (69% of those analyzed) responded in relation to allocentric coordinates, in many cases in relation to a local frame of reference and in some cases in relation to the absolute position of the stimulus in the room. In Experiment 2, the possibility of retinotopic encoding was investigated by presenting test spots of light at different positions relative to a fixation spot, and in different blocks of trials by moving the fixation spot to different positions on the screen. It was found that very few hippocampal cells were responsive in this task, and that for the cells that did respond, the encoding was not retinotopic. These results are consistent with the scarcity of egocentric encoding cells, and the preponderance of allocentric encoding, found in Experiment 1. In Experiment 3, it was found that relatively many hippocampal neurons (17%) responded differently according to the spatial position being fixated on the screen, in a task in which a small fixation spot appeared on each trial in a different position on the screen. This result confirms that hippocampal spatial cells do not encode spatial information using retinotopic coordinates, shows that it is sufficient for many of these hippocampal spatial neurons to respond that the monkey fixate particular positions in space, and is consistent with the finding in Experiment 1 that the preponderant type of encoding used by these hippocampal spatial neurons is allocentric.

Information processing in the temporal lobe visual cortical areas of macaques

Abstract: Visual pathways project by a number of cortico-cortieal stages from the primary visual cortex, the striate cortex, until they reach the inferior temporal visual cortex, area TE (Cowey, 1979; Gross, 1973; Desimone and Gross, 1979; Seltzer and Pandya, 1978; Rolls, 1990a,d; see Fig. 1). The inferior temporal cortex is among the highest unimodal visual cortical areas known. In addition, there are a set of cortical areas in the anterior part of the superior temporal sulcus which also receive visual inputs (Seltzer and Pandya, 1978; Baylis, Rolls and Leonard, 1987) (see Fig. 1). Of these areas, TPO receives inputs from temporal, parietal and occipital cortex; PGa and IPa from parietal and temporal cortex; and TS and TAa primarily from auditory areas (Seltzer and Pandya, 1978).