changeset 535:8a5abd51cd4f

add example / discussion per Winston's request.
author Robert McIntyre <rlm@mit.edu>
date Sun, 27 Apr 2014 20:25:22 -0400
parents 39ee58fef9d8
children 0b0fef5e817b
files thesis/cortex.org
diffstat 1 files changed, 48 insertions(+), 4 deletions(-) [+]
line wrap: on
line diff
     1.1 --- a/thesis/cortex.org	Sun Apr 27 10:17:45 2014 -0400
     1.2 +++ b/thesis/cortex.org	Sun Apr 27 20:25:22 2014 -0400
     1.3 @@ -2816,7 +2816,7 @@
     1.4     /empathy/ function will provide a bridge to use the body centered
     1.5     action predicates on video-like streams of information.
     1.6     
     1.7 -** Empathy is the process of tracing though \Phi-space 
     1.8 +** Empathy is the process of building paths in \Phi-space
     1.9  
    1.10     Here is the core of a basic empathy algorithm, starting with an
    1.11     experience vector:
    1.12 @@ -2845,10 +2845,54 @@
    1.13  
    1.14     Finally, to infer sensory data, I select the longest consecutive
    1.15     chain of experiences that threads through the sets of similar
    1.16 -   experiences. Consecutive experience means that the experiences
    1.17 -   appear next to each other in the experience vector.
    1.18 -
    1.19 +   experiences, starting with the current moment as a root and going
    1.20 +   backwards. Consecutive experience means that the experiences appear
    1.21 +   next to each other in the experience vector.
    1.22 +
    1.23 +   A stream of proprioceptive input might be:
    1.24     
    1.25 +   #+BEGIN_EXAMPLE
    1.26 +   [ flat, flat, flat, flat, flat, flat, lift-head ]
    1.27 +   #+END_EXAMPLE
    1.28 +
    1.29 +   The worm's previous experience of lying on the ground and lifting
    1.30 +   its head generates possible interpretations for each frame: 
    1.31 +
    1.32 +   #+BEGIN_EXAMPLE
    1.33 +   [ flat, flat, flat, flat, flat, flat, flat, lift-head ]
    1.34 +      1     1     1     1     1     1     1     4     
    1.35 +      2     2     2     2     2     2     2   
    1.36 +      3     3     3     3     3     3     3
    1.37 +      7     7     7     7     7     7     7
    1.38 +      8     8     8     8     8     8     8
    1.39 +      9     9     9     9     9     9     9
    1.40 +   #+END_EXAMPLE
    1.41 +   
    1.42 +   These interpretations suggest a new path through phi space:
    1.43 +   
    1.44 +   #+BEGIN_EXAMPLE
    1.45 +   [ flat, flat, flat, flat, flat, flat, flat, lift-head ]
    1.46 +      6     7     8     9     1     2     3     4
    1.47 +   #+END_EXAMPLE
    1.48 +
    1.49 +   The new path through \Phi-space is synthesized from two actual
    1.50 +   paths that the creature actually experiences, the "1-2-3-4" chain
    1.51 +   and the "6-7-8-9" chain. The "1-2-3-4" chain is necessary because
    1.52 +   it ends with the worm lifting its head. It originated from a short
    1.53 +   training session where the worm rested on the floor for a brief
    1.54 +   while and then raised its head. The "6-7-8-9" chain is part of a
    1.55 +   longer chain of inactivity where the worm simply rested on the
    1.56 +   floor without moving. It is preferred over a "1-2-3" chain (which
    1.57 +   also describes inactivity) because it is longer. The main ideas
    1.58 +   again:
    1.59 +   
    1.60 +   - Imagined \Phi-space paths are synthesized by looping and mixing
    1.61 +     previous experiences.
    1.62 +
    1.63 +   - Longer experience paths (less edits) are preferred.
    1.64 +
    1.65 +   - The present is more important than the past --- more recent
    1.66 +     events take precedence in interpretation.
    1.67  
    1.68     This algorithm has three advantages: 
    1.69