2016-03-04

Perception Test Run – 1, 4th March 2016

In this test run I have made every binon familiar upon its creation, made them ActState=Active and started a stem binon for them so all possible overlapping combinations are created and recognized. Thus it automatically creates all possible combinations at all the levels.

However at this point it is the first time a binon is recognized that already exists – binon #37 = 1xN1 and 1xN7 = “*T”.  However the previous stem binon was “E*” = binon #65 = 1xN9 and 1xN1. So what happened was the stem binon processing created a new binon #65 and then the recognition processing perceived binon #37. And binon#65 did not get combined / overlapped with binon #57. If ActStates were inactive then binons would have to be practiced before becoming familiar but I might still have the same problem.

The problem arises from the fact that stem binons have to wait for their goals to stop firing (that is an unexpected goal stimulus must occur), before they can be created. However existing binons fire as soon as their IDL ratio is reached.

Maybe existing binons should also stay perceiving until their goal binon stops repeating and then their IDL would be checked. Ones which have received more that their expected goal repeats could be idled but the others would be waiting for their goal to stop repeating.