The perception lobe works slightly differently to the other brain lobes so I've created this tutorial to demonstrate what makes it special.
The purpose of the perception lobe appears to be to summarise everything that a norn can 'perceive' into one brain lobe so the norn can eventually store concepts and memories based upon these perceptions (done in the concept lobe). This tutorial attempts to answer the questions about how the perceptible lobes are summarised into the perception lobe and how to work out what each cell of the perception lobe means.
The usual method of passing cell information from one lobe to another is through wiring up dendrites. Using dendrites you can only summarise information from two other brain lobes (the D0 and D1 dendrite connections). With the perception lobe there is more than two lobes that need to be summarised so it appears a mechanism specifically for dealing with the perception lobe was built.
In the genetics kit each brain lobe has an option that can be set indicating whether or not the lobes data is copied to the perception lobe.
As can been seen from the above dialog box from the general sense lobe the
options for perception are:
The brain lobes that are marked as 'Yes' or 'Mutually exclusive' in an original norn are:
| Lobe | Size | Data copied to perception lobe? |
|---|---|---|
| Drive lobe | 16 | Mutually exclusive |
| Verb lobe | 16 | Mutually exclusive |
| General sense lobe | 32 | Yes |
| Attention lobe | 40 | Yes |
The perception lobe must have a number of cells equal to or greater than the total number of cells in all lobes marked as 'Yes' or 'Mutually exclusive'. The total number of cells in the above lobes equals 104. The size of the perception lobe is 112 so there is a little room to spare there.
My theory from observation is that starting from the lowest numbered perceptible lobe, the output of each cell of that lobe is copied to the lowest available cell in the perception lobe. So with the perceptible lobes listed above I think the mapping to the perception lobe cells will be:
| Perception cell number | Other lobe cell number |
|---|---|
| 0-15 | Drive lobe 0-15 |
| 16-31 | Verb lobe 0-15 |
| 32-63 | General sense lobe 0-31 |
| 64-103 | Attention lobe 0-39 |
I do not know what the setting 'Mutually exclusive' means. From my tests it appears to do the same as 'Yes' but further experimentation may show otherwise. The following examples will attempt to demonstrate whether my theory about how the perception lobe works and is mapped is correct or not.
For this example we start with a norn with the normal nine lobes selected in Creatures and we will test the lobes with Perceptible marked as 'Yes'. Run the BrainCellMonitor program, connect to Creatures, and use the 'Add' button to view the following Lobe/Cell/Dendrites:
This will allow us to view the first and last cell in the general sense lobe and compare it against the perceptible cell we think they will be copied to.
We will now use CAOS commands to fire particular cells in the general sense lobe (lobe 5) and see if the corresponding cell in the perception lobe (lobe 0) fires with equivalent values.
Try executing the CAOS command: inst,trig 5 0 255,endm
Notice how the perception lobe cell number 32 output number increases and starts to decrease in a similar way to the general sense lobe cell we just fired:
This seems to imply that our assumed mapping may be correct.
Try executing the CAOS command: inst,trig 5 31 255,endm
Notice how the perception lobe cell number 63output number increases and starts to decrease in a similar way to the general sense lobe cell we just fired. This also seems to confirm our mapping. Try using different cell numbers with lobe 5 and you will see that the mapping we came up with is correct. Now we'll try it with the Attention lobe. Use BrainCellMonitor to view the following cells:
Try executing the CAOS command: inst,trig 7 0 255,endm
And
executing the CAOS command: inst,trig 7 39 255,endm
Once again you should see the equivalent perception lobes firing.
You may notice that when doing the General sense example that if you fire both general sense lobes then both perception lobes also fire. But in the Attention example, firing both will only cause one perception lobe cell to fire - why is this? It is because the Attention lobe is marked 'Winner takes all'. This means that only the cell with the highest output actually fires - all the others get automatically set to zero and this data is copied to the perception lobe.
Well what about the 'Mutually Exclusive' option? Lets try some examples with the Verb lobe which is marked "mutually exclusive". I don't use the Drive lobe just yet as the drives for a norn are constantly changing and can get in the way of our observations. Try viewing the following:
Try executing the CAOS command: inst,trig 3 0 255,endm
And
executing the CAOS command: inst,trig 3 15 255,endm
Once again you should see the equivalent perception lobes firing. The verb lobe is also marked 'Winner takes all' so you'll see similar effects to that shown in the Attention lobe.
You can try examples with the drive lobe as well but as mentioned before the norn drives can get in the way of observation.
The Mutually Exclusive option works in the same manner as lobes marked as 'Yes'. The difference between the two options appears when dendrites are linked between the perception lobe and the concept lobe. Usually from 1 to 3 perception cells are linked to a particular concept cell. The perception cells that are linked can change during the lifetime of a norn based on experiences it has encountered. But only one cell from a given Mutually Exclusive lobe can contribute to the forming of a particular concept. Drive is mutually exclusive so we can have a concept of 'hungry' but not 'hungry' and 'tired'. Verb lobe is also mutually exclusive so we can have a concept of being told to 'push' but not being told to 'push' and 'pull'.
I hope this tutorial/discussion was useful to you in examining the Perception lobe. If you come up with any interesting information I'd love to hear from you. I've updated the BrainActivity program to display the cell details in the perception lobe based on the above observations (as of version 1.3).