Animal intelligence

Part 1: Summarise the paper in (250 words).

The article focuses on the analysis of a spiking neuron model known as the spaun, which is composed of 25 million neurons (Chris et al. 1202). The aim of the article is to address the central challenge for the neuroscience of explaining the manner in which complex brain activity generate complex behaviour. This model can perform a number of behaviourally relevant functions, but this paper focuses on eight different tasks that this model can perform without any modification. Some of these tasks are not limited to reinforcement learning, serial working memory, and image recognition, among others. Initially, there have been numerous other neuron models but all of them had been focusing on larger number of neurons rather than how the simulations connect to various specific observable behaviours. The functional architecture of the Spaun model is composed an action selection mechanism, a three compression hierarchies, and five subsystems. The compression or decompression mechanism takes place in a natural way, just the same way neurons respond to internal or external stimuli. However, all the components of this model communicate using spiking neurons which are broadly referred as semantic pointers. The action selection mechanism is a process of higher dimensional neural representation where the basal ganglia determine the state in which the network should be in. In addition, each of the five subsystems has its unique role including information encoding, and transformation calculation.

However, the Spaun model has both advantages and disadvantages as far as demonstrating how large neuron activity contributes to a specific behaviour. For example, this model provides a distinct opportunity concerning how to test learning algorithms in a biologically related situation. Apparently, it is challenging for the model to show how the dynamic and complex neurological system develops from birth.

Part 2:

 Is the main goal of the study biology or engineering? Justify your answer [60 words]

The main goal of the study is more engineering than biological, and this is due to a myriad of factors. For example the entire Spaun model has been manufactured to represent how neurons transmit information from one another until when response to internal or external stimuli is achieved. Consecutively, the composition and functionality of the model demonstrates a high level of the numerous transmissions which occur in the engineering field.

Discuss how embodiment and situatedness affect the system developed. Do either of them simplify the computation performed? [up to 250 words]

The manner in which the Spaun model captures the general psychological characteristics of behaviour is greatly influenced by the way the model is situated. Precisely, if an item or behaviour to be recalled is situated at a far distance from the model, the latter is not likely to produce or recall the features of the item accurately, as opposed to the items that are placed closer to it. This means that the manner in which the items are placed closer to the model, or rather the manner in which the developer of this model places it closer to the items that needs to be recognized by the model simplifies its potential of recognizing them. Consecutively, the item must be placed in the dorsa lateral prefrontal context in order for the basal ganglia reroute cortical in order for the Spaun to recall it (Chris et al. 1203). Moreover, the embodiment of the Spaun model creates new hypotheses concerning the way various tasks are executed. For example, the proposed method of solving the rapid variable creation task is slow, compared to the way the human brain can create. In regard to this, I don’t think that the embodiment of the Spaun model simplifies the computation that it performs. For example, though this model facilitates the aspect of exploring the robust flexibility of the biological cognition, it fails to establish how learning takes place in the biological brains. Precisely, despite the Spaun model being complex to the extent that it can demonstrate the performance of 25 million neurons, it fails to explain how robust learning can take place in a continuously operating and multifunctional brain.

Suggest an extension that would give or enhance engineering performance. [60 words]

In order to effectively demonstrate the manner in which the basal ganglia manage to recall the psychological features of an object that is placed or stored in the dorsolateral prefrontal cortex of the Spaun model, I think an extended mode of object mapping needs to be embraced. For example, the objects need to be placed at a far distance than they have been placed in this study.

Suggest an extension or change that would give or enhance biological understanding. [60 words]

I think the performance of the Spaun model needs to be evaluated when a lower number of neurons are used and when a higher number of neurons have been used. This would enable the reader to establish whether or not there is a difference in terms of the functioning of the neurons or transmission of impulse when fewer neurons are involved.

Work Cited

Chris Eliasmith, Terrence C. Stewart, Travis Dewolf, and Daniel Rasmussen. A Large-Scale Model of the Functioning Brain.