Welcome to the Creatures Wiki! Log in and join the community.

Brain

From Creatures Wiki
(Redirected from Neurones)
Jump to navigation Jump to search

The brain in Creatures is a very complex thing built of about 10 lobes, maybe 900 neurons and many thousands of dendrites. The brain works with the creature's biochemistry in order to keep itself alive.

Lobes[edit]

A lobe is a part of the brain that is dedicated to a certain function. Each lobe has an x, y, width and height coordinate, giving it a unique position in the grid of the brain.

A lobe contains neurons, and, in Creatures 1 and Creatures 2, dendrites.

A common mutation of the brain is the Multi-lobed creature, with duplicate and/or new brain lobes. For articles on particular lobes, see the Brain category

Neurons[edit]

A neuron is a place where you can store a number value. Most neurons lose the saved value over time, but some do this faster than others.

Neurons live in lobes, and are connected to other neurons via dendrites.

Dendrites[edit]

Dendrites are the connections between different neurons. Dendrites work unidirectionally (one-way), so the value of neuron 1 may influence the value of neuron 2, but not vice versa. Dendrites may have different behaviours and some just transfer the value of one neuron to another one, while others may negate the value or do even more complex work. See also concept lobe.

In Creatures 1 and Creatures 2, dendrites live in lobes. After that, they live in tracts.

SVRules[edit]

Creature brains use a neuronal processing system called State Value Rules, commonly shortened to SVRules.

These were fairly simple in Creatures 1 and Creatures 2, essentially designed to let you perform simple manipulations upon Creature variables (ie, biochemistry and brain values), with the possibility of adding conditions based upon these variables.

In Creatures Evolution Engine games such as Creatures 3, each neuron and dendrite is a fully functional register machine. That means that you have several registers (places where you can store something) and one special working register. The most important commands on a register machine are LOAD, which loads a value into the working register and STORE, which stores a value from it into one of the other registers.

Operations like ADD always use the value of the working register. Look at the following example, which adds the value from register 1 to the one of register 0 and stores the new value in register 2:

  • LOAD 0
  • ADD 1
  • STORE 2

The SVRules system adds some extensions to the usual register machine, like access to the Creature's chemical system or to its reception.

Take your favourite genetic editor and take a look into one of the Brain Lobe or Brain Tract genes to get a full (?) list of supported operations.

A good place to learn more about the brain is The Creatures Developer Resource

A note from the programmer[edit]

This SVRule system was designed to deal with an interesting problem: How could I specify arbitrary behaviours for my neurons in a way that evolution could freely change, without it generating endless syntax errors in the process? For instance, imagine that a mutation were to occur in a line of conventional C++ code, such as "for (i=0; i<num; i++)". Almost every possible mutation would render the code unreadable (e.g. "fxr (i-0; ibnum; i4+)") and the chances of a useful (or even viable) mutation would be extremely low. At worst the whole simulation would crash.

To solve this problem I designed the SVRule system in such a way that EVERY statement you can write in it is legal and meaningful, even if it is not biologically useful. If a token was originally the operand for a command, for example, and the command later mutated to one that didn't require operands, the token would now be interpreted instead as a new command or a variable. The details of this aren't very important, but I guess it's an interesting example of the ways in which biology differs from computer programming. Biology tends to have this sort of robustness built into the design.

Digitalgod 12:39, 4 Apr 2005 (EDT)

Brain Organ[edit]

In the games after C2, one way in which creatures can die is by the lifeforce of the brain organ - which contains only the brain lobes - becoming too low. This happens when the organ receives sufficient damage which, due to the lack of other biochemistry inside the organ, only happens when a creature has insufficient ATP. The gene which governs this organ has a similar structure to other organ genes.

Tracts[edit]

Tracts were introduced in the Creatures Evolution Engine as a new object for managing dendrites. Each tract defines a set of dendrites connecting between two lobes. This allowed brains to be more complex compared to Creatures 1 and Creatures 2, in which each lobe could only have up to two types of dendrites connecting to it.

Implementation details[edit]

C1 lobes[edit]

The lobes in a Creatures 1 norn are as follows (this also applies to most of the other Cyberlife/Creature Labs-provided creatures):

Number Name X Y Width Height Neurones
0 Perception 4 13 7 16 112
1 Drive 34 30 8 2 16
2 Source 15 24 8 5 40
3 Verb 37 24 8 2 16
4 Noun 21 3 20 2 40
5 General Sense 32 34 8 4 32
6 Decision 53 15 1 16 16
7 Attention 44 30 5 8 40
8 Concept 12 6 40 16 640

C1/C2 SVRules[edit]

C1 opcode C2 opcode name description
0 0 <end> Marks the end of an SVRule. Any opcodes appearing after this marker are ignored.
1 1 0 The integer number zero.
2 2 1 The integer number one.
3 3 64 The integer number sixty-four.
4 4 255 The integer number two hundred fifty-five.
5 5 chem 0 Current amount of chemical 0 in the brain lobe. This chemical can be changed with a biochemical receptor.
6 6 chem 1 See above.
7 7 chem 2 See above.
8 8 chem 3 See above.
9 9 state Current value of the state of the cell.
10 10 output Current output value of the cell.
11 11 thres Current threshold of the cell.
12 12 type 0 The sum of type 0 dendrites. The individual value of each dendrite for this type is summed and returned. The individual value of a dendrite appears to be calculated as: value = source cell * (stw/255) where 'source cell' is the value of the cell that this dendrite is attached to from the source lobe and 'stw' is the dendrites current short term weight value.
13 13 type 1 See above.
14 14 anded 0 If all type 0 dendrites are firing then this will be the value of the sum of these dendrites. If any of the type 0 dendrites are not firing then this value will be 0.
15 15 anded 1 See above.
16 16 input Current dendrite input signal.
17 17 conduct Not hooked up in Creatures 1. Unknown in Creatures 2.
18 18 suscept Current susceptibility to reinforcement. This is the value calculated by the 'susceptibility' state variable rule on the dendrite dynamics pages of the genetics kit. This opcode should only appear in the svrules associated with dendrites (otherwise there is no way to determine what dendrite it is pulling the data from).
19 19 STW The value of the short term weight setting for the given dendrite. This opcode should only appear in the svrules associated with dendrites (otherwise there is no way to determine what dendrite it is pulling the data from). The short term weight for a dendrite appears to be calulated using the following formula with a relaxation decay applied to it: stw = ltw + (susceptibility/255) * reinforcement
20 20 LTW The value of the long term weight setting for the given dendrite. The long term weight acts as a rest state for short term weight. STW and LTW reduce towards each other with LTW moving slower than STW. This opcode should only appear in the svrules associated with dendrites (otherwise there is no way to determine what dendrite it is pulling the data from).
21 21 strength Current value of dendrite strength. This opcode should only appear in the svrules associated with dendrites (otherwise there is no way to determine what dendrite it is pulling the data from).
22 32 The integer number thirty-two.
23 128 The integer number one hundred twenty-eight.
24 rnd const
25 chem 4 Current amount of chemical 0 in the brain lobe. This chemical can be changed with a biochemical receptor.
26 chem 5 See above.
27 leak in This appears to be related to the value of the 'back propagation' state variable rule. If any other lobe has this current lobe as a 'source lobe' in its dendrite settings then the value of 'leak in' for the current lobe is the value of the 'back propagation' svrule from the destination dendrite. This can be used to feed data back to a source lobe from destination lobes.
28 leak out This appears to be related to the value of the 'forward propagation' state variable rule. If the current dendrite has a 'forward propagation' svrule then the 'leak out' opcode is the value of that 'forward propagation' calculation.
29 curr src leak in The neuron's current dendrite's source neuron leak in charge.
22 30 TRUE If the previous opcode evaluates to TRUE (ie. not zero) then execute the remaining opcodes otherwise the value of state is zero and the SVRule completes.
23 31 PLUS Adds the value of the following opcode to the value of the previous opcode.
24 32 MINUS Subtracts the value of the following opcode from the value of the previous opcode. For example state:MINUS:1 will subtract '1' from the current value of 'state'.
25 33 TIMES Takes the left hand opcode multiplied by the right hand opcode and divides this by 256. The result of this calculation is the value of this opcode. For example: 64:TIMES:thres where 'thres' is 32 will be 32*64/256=8.
26 34 INCR Returns the value of the previous opcode incremented by one.
27 35 DECR Returns the value of the previous opcode decremented by one.
36 FALSE If the previous opcode evaluates to FALSE (ie. zero) then execute the remaining opcodes otherwise the value of state is zero and the SVRule completes.
37 multiply Multiplies the value of the following opcode with the value of the previous opcode.
38 average Averages the value of the following opcode with the value of the previous opcode.
39 move twrds
40 random This opcode takes two arguments. It returns a random number between the first argument and the second argument. It also sets rnd const.

C3/DS SVRules[edit]

Opcodes[edit]

0: stop Stops processing of the current rule.
1: blank Set the specified destination to zero.
2: store in Copy the contents of the accumulator to the specified destination. When storing to a neuron value, The Brain-in-a-vat seems clamped to [-1.0, 1.0]. To check.
3: load from Copy the contents of the specified distination to the accumulator.
4-15: if commands Only execute the following opcode if the specified condition for the accumulator is true.
16-21: basic arithmetic Perform the specified arithmetic operation on the accumulator.
22-23: minimum/maximum with Sets the accumulator to the minimum/maximum value of the current accumulator value and the specified value.
24: set tend rate Set the rate (as a fraction of the difference) at which to 'tend' towards things using the 'tend to' opcode. Zero by default. Negative values seem to have the same effect as positive ones.
25: tend to Modify the accumulator so it moves further towards the operand, using the fraction specified in 'set tend rate' to decide how much of the difference to add/remove from the accumulator.
26-27: load negation/abs of Set the accumulator to the negation or absolute value of the operand. For instance, loading the negation of 0.5 will result in -0.5 and vice versa. However, loading the absolute value of either 0.5 or -0.5 will result in 0.5.
28: get distance to Set the accumulator to the absolute value of the difference between the accumulator and the operand.
29: flip around Subtracts the accumulator from the operand, and stores the result in the accumulator (ie, same as 'subtract from').
30: no operation Do nothing.
31: register as spare Register this neuron(?) as spare. Used by: decision lobe, attention lobe
32-33: bound in range Bind the operand between (inclusively) either [0,1] or [-1,1], and store the result in the accumulator. For instance, binding 1.4 to [0,1] would result in 1.0 being stored, but binding -0.2 to [-1,1] would result in -0.2 being stored.
34-35: add/tend to and store in Like either the 'add' or 'tend to' opcodes, but store the result in the operand instead of the accumulator.
36: Nominal Threshold If the accumulator is less than the value of the operand, set the accumulator to zero. Otherwise, do nothing.
37: Leakage Rate Set the leakage rate (used by rest state) to the value of the operand.
38: Rest State ??
39: Input Gain Hi Lo ??
40: Persistence drives neuron State towards neuron Input value: S(1) = S(0) + ( (1-Persistence) * (I(0) - S(0)) )
41: Signal Noise ??
42: Winner Takes All ??
43: ShortTerm Relax Rate Set the short-term relaxation rate. Used by: drive to combination tract
44: LongTerm Relax Rate Set the long-term relaxation rate, and does the relaxation calculation. Off the top of fuzzie's head, it goes like this (hardcoded vars): weight = weight + (Var 1 - weight) * STR; Var 1 = Var 1 + (weight - Var 1) * LTR ie, it's intended to pull the weight downwards towards Var 1 (stable state) at a speed according to STR, and Var 1 upwards towards the weight at a speed according to LTR.
45: store abs in ??
46-47: stop if zero/nonzero ??
48-49: if zero/nonzero goto ??
50-51: divide/multiply by, add to neuron input This seems to, indeed, divide or multiply by the operand, and add to the neuron Input variable (var 1).
52: goto line ??
53-56: stop if condition ??
57: Reward Threshold ??
58: Reward Rate ??
59: Use Reward With ??
60: Punish Threshold ??
61: Punish Rate ??
62: Use Punish With ??
63: Preserve Neuron SV Very weird. If I remember correctly, it basically stores the value of the neuron variable given by the *value* of the operand, in the neuron Var 4. ie, if you pass it a value '3.0' it will take Var 3 and store that in Var 4.
64: Restore Neuron SV I think this does the opposite of 63.
65: Preserve Spare Neuron Like 63 (preserve neuron SV), only with the spare neuron.
66: Restore Spare Neuron I think this does the opposite of 65..
67-68: if negative/position goto ??

Operand Types[edit]

0: accumulator The accumulator. Useful for instructions which only operate on an operand, such as the "bound in" opcodes.
1: input neuron The source neuron for the dendrite which this svrule is running on?
2: dendrite The dendrite which this svrule is running on?
3: neuron The neuron which this svrule is running on? (or the destination neuron, for dendrites)
4: spare neuron ??
5: random A random number between 0.0 and 1.0.
6: source chemical ??
7: chemical A chemical level (with chemical# specified)?
8: destination chemical ??
9: zero The literal value 0.0.
10: one The literal value 1.0.
11: value A literal value from 0.0 to 1.0.
12: negative value A literal value from 0.0 to -1.0.
13: value x 10 A literal value from 0.0 to 10.0.
14: value / 10 A literal value from 0.0 to 0.1.
15: value integer A literal value from 0 to 248.

External links[edit]