all-in allows the neuron to find a pattern in all of its transpose positions. all-in is activated by giving the in-slot, pattern and from where the transpose levels begin and end (from 0 to 3 in the example). The position value is given as usual. This neuron finds patterns (a b a) (b c b) (c d c) from the input pattern.
(def-neuron test-pattern5
(all-in 1 '(a b a) 0 3 0) 'x
(otherwise '=)
)
(run-neuron 'test-pattern5
'(a b a b c b c d c))
--> (x = = x = = x = =)
The following finds the same patterns that exist in shifted form in both the in 1 and in 2 patterns.