Modeling the Neural Network to Test the Basis of Working Memory in Chimpanzee

Abstract

Visual-spatial memory is essential for forming a representation of whatever picture the eye captures in mind. Working memory is a subset of the limited capacity part of memory that helps in cognition by integrating information modulation and transient storage (Baddeley & Hitch, 1974). The working memory makes some amount of information to be readily accessible by retention of that piece of information. A study at the Kyoto Primate Research Institute in s series of experiments showed that Ayumu, the chimpanzee they have been training for various task-based purposes, has performed better at memory-based tasks than their human counterparts (Inoue & Matsuzawa, 2007). Non-human primates are recognized for their exceptional ability to process local characteristics, whereas humans excel in integrating the local details to form a comprehensive global perception (Imura & Tomonaga, 2013). As much as visual temporal integration plays a role here, the integrated effect of attention and working memory to reproduce temporal cues is also equally important (Marchetti, 2014). Given the environment that the chimpanzees and humans live in are wildly different, the visual processing system may also have evolved some specialized structures. In order to gain a better understanding of the complex networks within neural systems, we employ models that are capable of reconstructing network topology from various forms of data - such as neural firing data. However, it is not enough to simply rely on these models alone. It is essential that we assess their stability before generating any further data through simulation in order to obtain accurate and reliable reconstructed network parameters for analysis. Through this process, we hope to achieve greater depth and precision in our understanding of the intricate workings of these highly complex networks.