ForF lobe notes (recap)

Vampess · #1 26-06-2007, 02:25 AM

Some of you might remember the ForF lobe notes thread from before the crash.

Ratboy posted about some oddities in the forf lobe, and the notes going with them. After some testing, it appeared that processing of instincts resets the State variable briefly in creatures.

While the general memory of a creature is managed by dendrites, the forf lobe is dependent on neurons, and more specifically on the State variable. Ratboy made a lobe that stores the State variable of the forf lobe, and restores it after instinct processing.

The creatures however, were still hardly reacting, or even oppositely. Looking into the forf lobe, I came to the following conclusion:

Quote:

I've studied the forf lobe, and how forf works, and it appears the drive stracts are more important than thought.

The most important variable is Input, it is derived from the drive tracts, and determines how susceptible the creature is to changes it's opinion. It's the same in each neuron, and at the moment is only derived from being lonely or hurt.

The stims associated with tickles and slaps, reduce or raise loneliness, fear, and anger. So when a creature is slapped or tickled, these drives are adjusted accordingly, which in turn determine how susceptible a creature is.

How does the creature get susceptible? The Susceptibility variable is set to 1 in the appropriate neuron by the forf command, and then decays rather quickly, so if it's pushed by norn A, and subsequently hit by norn B, it get's a positive reinforcement with norn A, and a negative with norn B.

The Output variable is used for dendrite migration between the forf and comb, and is derived from the mood lobe, in other words; the mood in which the creature is in, will affect what they do towards other creatures.

This works pretty good, except for the fact that 2 of the 4 tracts are currently working, so the creature is actually getting one third of the feedback it's supposed to get. This way, it decays faster than that it gets feedback.

After adjusting the the tracts, it was more consistent, but passing into Youth, their behaviour became erratic again. Investigating the responsible tract, I adjusted the following:

Quote:

Here's a list of what I've done, sorry for not posting it earlier but I was too excited to see it work.

1. The drive tracts mentioned in the first post, I fixed the scared, and the angry/friendly tracts to match the hurt and lonely tracts.

The reason I did this was because these tracts provide the input for an opinion, after disabling all the tracts, the creatures got no input, or opinions at all. But after fixing them, it was still oddly going down. And convinced as I was that those tracts should be working, I continued on testing with these tracts intact.

After a while I noticed one similarity with all the norns; their opinion would stay pretty stable up until they reached Youth. I remembered looking extensively at the reproduction genes when making my own norn, and I remembered: Females' friendliness only goes up and down in cycles, at which stage, she just starts pushing more in general. A males' friendliness only rises when they're pushed (and then just decays if not pushed again).

But looking into the stimulus, I found out that being pushed has no effect on friendliness, but the mating stimulus does. It sharply reduces friendliness in both norns.

The way the tracts are set up, creatures get a good opinion of another when the action lowers these drives, and a bad one if it raises them. So when a female norn is on her cycle, and furiously pushes a male norn, and he pushes back, but they're not conceiving, they'll start to dislike eachother after a while.

First off, the angry and friendly drives were stuck together (just because they're next to eachother I guess), so I cut it in half, leaving the angry part, and made another tract for friendliness. Just turning around the tract wouldn't work, as norns would slowly start to like eachother, but dislike eachother immediately after conceiving.

2. So I made two parts to the tract, and tied it to a chemical closely related to sex drive; libido lowerer. The advantage is that because younger norns don't have any reproductive chemical, it works just the same for them. The other part is that when fertile norns do conceive, they'll both take a higher liking for eachother than otherwise. Which I think is pretty realistic.

Soon I had alot of loving and promiscuous norns wandering around, seeing as they'd like eachother from the start, and be prone to push eachother, everyone was very loved, very soon.

After tweaking it some, I found out that this was definitely the best way to go.

Somehow, my attention wandered off to meeting new hands, I found out that the support lobe overwrote it, making the opinion for the current hand, the opinion for all hands. That was a feature I didn't want to lose.

3. I adjusted the support lobe to reset the neurons when Susceptibility Input is active, just as the regular lobe would do.

After testing this with a single norn, I put a couple of more in the meso, intending to continue my search to, believe it or not, make creature meet neutral, so there'd be less default loving, and more suspicion towards grendels, seeing as creatures first tried to push them, before getting beaten.

To my suprise however, the Susceptibility Input isn't only set when meeting new hands (although it's the only one set on 0), but also when meeting new creatures. The result was it that it overwrote the initial value, but stored the rest.

This does result in more mischievous creatures, especially when they're born, and have high drives. The reason is simply the fact that they're not standard prone to pushing another, but rather act on mood and memory. And ofcourse you can reinforce memory by training them.

I tweaked the friendliness tract some more, and found out that the mood's lonely and friendly tracts were off, so I adjusted them to multiply, which proved to be the best solution to provide more impact.

Last off I found out that the hurt -> forf tract had 15 connections instead of 1 (the rest is useless), and I adjusted the lobe ID to the only free slot under 10, 4.

So, I have all official breeds ready, including the original grendel & ettin, which are quite a story on their own, but I might post about that later.

But I want more feedback, because I want to get the kinks out while they're still in beta phase, that's why I'm only making the ChiChi* version public

* Based on Ratboy's genome.

malkin · #2 26-06-2007, 03:12 AM

So these norns will be more sociable in general? I'd hate to think of what'd happen if these genes got into the Shagadelics...

I've compiled the genome into an egg agent with EasyGMS - I left all the settings as default except for the colour - they're a cute burgundy now.

Vampess · #3 26-06-2007, 03:36 AM

Uhm, malkin, these guys are beta phase, I'll make eggfiles when they're done.

malkin · #4 26-06-2007, 04:05 AM

I guessed that - but it's difficult for me and other people who use macs to inject a plain genome into the game. I made the egg file of the beta so that I could inject them easily, and then I thought I'd put it up for other people who have macs. No sense reinventing the wheel and whatnot.

Vampess · #5 26-06-2007, 04:16 AM

Should've known.. In the old thread I PM'ed Embri the code to import them through CCL. But yeah, I understand, I just really want experiences with the breed, my apologies

The reason I don't want beta eggfiles floating around is because they're subject to change frequently, so it's a timewaste, and pretty confusing for people testing them.

malkin · #6 27-06-2007, 12:49 AM

It's difficult to know what they want because they're always going 'eem likes Ron' whenever you ask them to express. A couple of them dislike each other, and I'm not fully sure why. From the speech bubbles, it *seems* to be a love triangle, but that just might be me anthromorphising. It's useful to have the Air Purifier injected, because they don't really seem to retreat norn when crowded... Can you have a go at a gene so that they become less crowded when they're sleeping?

Vampess · #7 28-06-2007, 01:45 PM

There is no normal gene modified compared to the original Chi-chi's genome, this is only a brain edit, anything to do with retreating while crowded has nothing to do with the ForF lobe, nor do they behave different in that aspect from the original genome. Perhaps you would like to read the first post again.

As for waiting your turn to ask something, is simply the queue of things to be processed. If you want to know how a specific norn feels, you'll need to single it out for a few seconds. Nothing I can do there, they already are alot less chatty as it is.

And as for anthropomorphising, that's what Creatures' appeal is based on. They look so real you'd almost think they are, even the brain model is based on the human's, in a very crude way, creatures make decisions in the same way we do.

Creatures have character, which never really shone through before. Due to all the different memories, and different chemistries, some creatures are more prone to be loved, other are more prone to be hated, or not cared about at all, even within subspecies.

With ettins this is most noticeable, as they get afraid and crowded from hearing other creatures. Some ettins get to mate, and eventually love eachother, while others go absolutely mad, and kill everybody (which is due to the ForF lobe, because being afraid of a creature triggers either retreat or hit). Well, that actually only happend once, so it's not like every bunch of ettins will contain a killer, but it's a possibility.

When I put Fiona in a bunch of edited Chi-chis, the result was amazing, at first, norns starting agreeing they didn't like Fiona, then 10 norns just jumped on her. Fiona even hit while retreating, and seemed pretty invulnerable, when the group got near the heating pad, I saw a child norn, who was sort of stuck at the heating pad, because everytime he got near Fiona, he retreated from her. One for one they fell, and I thought it would be over in minutes, but a while later, there were 2 baby males and an adult female on the bottom level, and Fiona at the top. One baby male ventured to the second level, and was promptly killed by Fiona. The other two stayed downstairs, the adult female disliked Fiona, having endured a beating before, while the now youth male never met Fiona.

The youth male, dared to go to the top level, and amazingly, wasn't killed, although Fiona was old already. A few minutes later, I saw the pregnant icon, expecting the male went down again (where the adult female still was), I found out he actually impregnated Fiona, and liked her. Not long after, Fiona found the female downstairs, and killed her. The offspring had the physique and temper of their mom, but while Fiona was alive, the male norn was too, and he still liked Fiona. After Fiona died, the male norn was killed by his son. The reason he was kept alive while Fiona was alive, was because he liked Fiona, so he stayed around her, and for some reason, Fiona liked hitting her offspring more then she did hitting him.

malkin · #8 04-07-2007, 12:22 AM

The 'become less crowded when sleeping' gene was a suggestion, I didn't think there was a gene edit in there already. I find that they become more crowded than my other creatures, so it might be worthwhile giving them a little something to help them cope with crowdedness.

It really seems to give them some character, disliking and liking and loving other creatures.

Vampess · #9 04-07-2007, 03:33 AM

I understand. But the point of beta testing these guys is to spot oddities in opinions, unless, ofcourse, you feel they are more crowded than regular Chi-chis.

If you, or anyone would like to base breeds on these guys when they're finished, that would be great, but my intention is to fix the ForF lobe. I've spent a good time looking at them while tweaking, and after that. The most I've been looking at however, is numbers, and that's what I see; I know when they should say what.

And that's why I need objective testers, to know if people feel comfortable -or even familiar- with how creatures form opinions. Don't get me wrong, I like them, but I want to have them double checked, before they go public.

malkin · #10 16-07-2007, 02:57 PM

I don't feel they're more crowded than normal, it's that (in gen 1 and 2) they don't seem to do anything sensible when they're crowded, like retreat, or push portal. I know that sensible norns are a tall order, lol. I think they develop the 'crowded does stuff all' response quicker than normal chichis. (gen 4)