Brain Simulator II : The Guide for Creating Artificial General Intelligence by Simon Charles

Author:Simon, Charles [Simon, Charles]
Language: eng
Format: epub
Publisher: FutureAI
Published: 2021-03-16T16:00:00+00:00

The Neuron Engine menu allows you to set the various engine parameters.

Editing Networks

Neurons

Clicking a neuron will cause it to fire. If it was firing continuously, clicking the neuron will cause it to stop firing.

Right-click a neuron to display its context menu. Each neuron has a numeric ID, which is the location of the neuron within the neuron array.

Right-click any neuron to display its context menu. Sub-menus show incoming and outgoing synapses.

The ID is followed by an optional label which you can add or edit. If you add a label, it label can be just a few characters and will reside within the neuron disk or it can be longer, extending beyond the disk of the neuron to act as a notation in the network. Neuron labels are not typically used in computation but can be used to reference the neuron. Duplicate labels are allowed but a warning is displayed if you set a label which occurs elsewhere in the network. The neuron label may be used when editing a synapse or when pasting a selection as described later.

In the dropdown, you can select the model to be used for this neuron. This is followed by the neuron’s charge or membrane potential. You can edit it. For this and the following entries, numeric text entries must be syntactically correct. Illegal entries will be ignored. Below this, several model-specific parameters may be displayed which are described at the end of this section. In this case, the Leak Rate and Axon Delay are specific to the LIF model.

The checkbox, “Record Firing History” will begin recording for this neuron and will open the firing history window if it was not already open.

Lists of synapses to and from this neuron are also available. Each list shows the weight and the target neuron (or source neuron in the case of an incoming synapse). The neuron’s label will be displayed if it has one. If you left-click a synapse entry, the context menu for the target (or source) neuron will be opened; the neuron need not be visible in the current display area. If you right-click the synapse entry, the context menu of the synapse will be opened.

The entry “Paste Here” will insert the content of the clipboard at this current location. The entry “Move Here” will move the content of the current selection to this current location. These are covered in detail under “Clipboard” below.

The “Connect Multiple Synapses” command has a submenu with three commands which act in concert with a selection. The command “Selection to Here” and “Here to Selection” will add a synapse with the current default characteristics between the current neuron and every neuron in the selection. The two commands differ in the direction of the synapses. The command “Mutual Suppression” will add a synapse of fixed weight -1 between every two synapses in the current selection.

For the following commands: If the neuron being changed is part of a selection (see “Clipboard” below), the change will apply to all the neurons in the selection:

Setting the neuron’s charge.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.