Computational modeling of pair-association memory in inferior temporal cortex.

摘要: Distinctive neuronal activities related to visual stimulus-stimulus association have been found in the inferior temporal (IT) cortex of monkeys. They provide an important clue elucidating memory mechanisms brain, but do not accord with existing neural network models. In present paper, we clarify computational principle required for reproducing empirical data and construct a biologically feasible model that learns performs delayed pair-association task. This is composed two networks, N1 trainer N2, memories are formed by their interactions. Specifically, N2 receives output addition external input, sends learning signal back N1; this works as guide shifts pattern or state transitions N1, traces engraved along its path, so trajectory attractor connecting from cue-coding target-coding N1. Computer simulation shows only distinguishes target task, also explains activity IT neurons very well. It reasonable presume correspond area TE rhinal cortex, respectively; based on theory, explain some physiological findings memory, make several predictions.