Adaptron Test Run – 1 - Date: 4th Nov 2015
I am repeating the = = = N sequence using the latest S-Habit activation state machine algorithm. Whenever the attended to stimulus changes the S-habits are formed from the last two previously attended to stimuli. At this point in the process the N = = = S-Habit#2 is expecting its goal stimulus (the 3rd = ) and it gets it. This adds S-Habit#2 (M2) to the perception list and it is now novel. It attracts attention. But this causes the formation of the S-Habit#4 (N = = ) because that was what was in the previously attended to stimuli. This is wrong. The code checks to see if the previous attended to stimulus is the same or different from the current attended to stimulus to make the decision about forming a brand new S-Habit. It should be checking to see if the previously attended to stimulus has repeated or not to make this decision. It should not compare it with the current attended to stimulus.
But when looking at the code it may be better to use some logic that says since the attended to stimulus is brand new and distracting the previous stimuli should not be combined into an S-habit.