[HTM Forum] [HTM Theory] Questions on displacement cells activation

2019-02-14 Thread Matt Taylor via HTM Forum


[quote="nick, post:1, topic:5439"]
Can there be displacement cells representing 2,3,n steps?
[/quote]

The example you referred to is a bit contrived because it is so small. In a 
real space, there might be millions of distinct places you might represent, 
each one of which could have a displacement applied. But yes, you could have a 
displacement that represented a jump from any location in space to any other. 
The steps don't really matter. Now think about this in a high-dimensional space 
and :exploding_head: 

I don't quite understand your 2nd two questions, but does my answer help 
explain anything?





---
[Visit 
Topic](https://discourse.numenta.org/t/questions-on-displacement-cells-activation/5439/4)
 or reply to this email to respond.

You are receiving this because you enabled mailing list mode.

To unsubscribe from these emails, [click 
here](https://discourse.numenta.org/email/unsubscribe/321b79ec54da53cf9fde7980f456165471b522baa62f6fed859ee886935b058a).


[HTM Forum] [HTM Theory] Questions on displacement cells activation

2019-02-14 Thread Matt Taylor via HTM Forum


Nick, I think I follow you up until the animations. I don't understand how 
either animation relates to the topic of displacement cells.





---
[Visit 
Topic](https://discourse.numenta.org/t/questions-on-displacement-cells-activation/5439/2)
 or reply to this email to respond.

You are receiving this because you enabled mailing list mode.

To unsubscribe from these emails, [click 
here](https://discourse.numenta.org/email/unsubscribe/b62bd5c74e4f763fc551dc3b5f8f21541a94d01f0ae3e0cb94c1afefc9db24f6).


[HTM Forum] [HTM Theory/Neuroscience] Questions on displacement cells activation

2019-02-14 Thread nick via HTM Forum


> For example, if cell G(1,1) is active for a location on an object, a movement 
> one step to the right will cause G(2,1) to become active. Since the cells 
> tile space, an additional movement one step to the right will cause G(0,1) to 
> become active.

![Figure%20A|388x465](upload://ouT6j3mB7WA7geIIwIcrJNOSMDa.png) 

The supplementary material covers single step movements. In the Grid cell 
module in the **Figure A** when the cell G(1,1) is active a single step 
movement can be 1 step to the "left", "right", "up", "down", "left-up", 
"left-down", "right-up", "right-down" (diagonally) or "none" (same place but 1 
step through time). All these count for 9 relative displacements of 1 step. 
That's why there are 9 Displacement cells representing them in the Displacement 
module.

![Figure%20B|307x451](upload://3vdkUpG6wztSYp8ECRhRNo9MXsF.png) 

**Figure B** shows that independent of which cell performs the 1 step to the 
right movement D(1,0) will represent the displacement. The formula in the 
supplementary material for calculating which displacement cell will become 
active is:

G(x1,y1) --> G(x2,y2): D((x2-x1) mod 3, (y2-y1) mod 3)

Testing with 1 step to the right:

G(0,1) --> G(1,1): D(1 mod 3, 0 mod 3) = D(1,0)
G(1,1) --> G(2,1): D(1 mod 3, 0 mod 3) = D(1,0)
G(2,1) --> G(0,1): D(-2 mod 3, 0 mod 3) = D(1,0)

Testing with 2 steps to the right:

G(0,1) --> G(2,1): D(2 mod 3, 0 mod 3) = D(2,0)

But D(2,0) is representing 1 step to the left.

>From G(0,1) 1 step to the left or 2 steps to the right will both land on 
>G(2,1) but the displacement is different.

Questions:
1. Can there be displacement cells representing 2,3,n steps?
2. Is this problem solved more elegantly by combining grid+displacement cells 
made to work on a different scale?
4. Is it done by extending the capabilities of displacement cells by changing 
the anatomy to different types of cells like conjunctive cells?





---
[Visit 
Topic](https://discourse.numenta.org/t/questions-on-displacement-cells-activation/5439/1)
 or reply to this email to respond.

You are receiving this because you enabled mailing list mode.

To unsubscribe from these emails, [click 
here](https://discourse.numenta.org/email/unsubscribe/8b39aba87bc8e8a5ab43027024f608ec20ff3c47902c64dcfdcce0cb3d404b71).


Subscription probe for IBM-MAIN - please ignore

2019-02-14 Thread The University of Alabama LISTSERV Server (16.0)


Thu, 14 Feb 2019 06:00:01

This message is a "probe" for your subscription to the IBM-MAIN list. You
do not  need to  take any  action to  remain subscribed  to the  list. In
particular, you should not reply to  this message. Simply discard it now,
or  read on  if  you would  like  to  know more  about  how this  probing
mechanism works.

A  "probe"  is a  message  like  the one  you  are  reading, sent  to  an
individual subscriber  and tagged  with a  special signature  to uniquely
identify  this  particular subscriber  (you  may  not see  the  signature
because it is in the mail  headers). If the subscriber's email address is
no longer  valid, then the message  will be returned to  LISTSERV and the
faulty address will be removed from the list. If the subscriber's address
is still valid, then the message will not bounce and the user will not be
deleted.