1.1 --- a/thesis/cortex.org	Sun Apr 27 21:50:25 2014 -0400
     1.2 +++ b/thesis/cortex.org	Sun Apr 27 21:52:39 2014 -0400
     1.3 @@ -3216,15 +3216,17 @@
     1.4  ** COMMENT Digression: Learning touch sensor layout through free play
     1.5  
     1.6     In the previous section I showed how to compute actions in terms of
     1.7 -   body-centered predicates which relied on the average touch
     1.8 -   activation of pre-defined regions of the worm's skin. What if,
     1.9 -   instead of receiving touch pre-grouped into the six faces of each
    1.10 -   worm segment, the true topology of the worm's skin was unknown?
    1.11 -   This is more similar to how a nerve fiber bundle might be
    1.12 -   arranged. While two fibers that are close in a nerve bundle /might/
    1.13 -   correspond to two touch sensors that are close together on the
    1.14 -   skin, the process of taking a complicated surface and forcing it
    1.15 -   into essentially a circle requires some cuts and rearrangements.
    1.16 +   body-centered predicates, but some of those predicates relied on
    1.17 +   the average touch activation of pre-defined regions of the worm's
    1.18 +   skin. What if, instead of receiving touch pre-grouped into the six
    1.19 +   faces of each worm segment, the true topology of the worm's skin
    1.20 +   was unknown? This is more similar to how a nerve fiber bundle might
    1.21 +   be arranged inside an animal. While two fibers that are close in a
    1.22 +   nerve bundle /might/ correspond to two touch sensors that are close
    1.23 +   together on the skin, the process of taking a complicated surface
    1.24 +   and forcing it into essentially a circle requires that some regions
    1.25 +   of skin that are close together in the animal end up far apart in
    1.26 +   the nerve bundle.
    1.27     
    1.28     In this section I show how to automatically learn the skin-topology of
    1.29     a worm segment by free exploration. As the worm rolls around on the