oh ya separate this ->

heybroicanseeyouaremrjojotheponyontheranchhahaseethatyesyouareaspaceape

My tree will store representations ex. word>word>word (using byte pair encoding 
segments), I'll use bytes to store a node ex. g>r>8>X>k>... is "wh en will the 
bo ok be read"......so when my AI will see "whenwillthebookberead", the wh 
activates the wh node, then ' en' activates the next part 'en' with a delay 
match ability to ignore the space but with discount - already working in my 
code, just I don't yet store such BPE segments nor online in a tree nor have 
yet the text that is original ex. 'wh en' since my preprocessor i use from 
Byron's gives me ex. 'Gf8k' to read/predict for. So I'l not only match using 
delay 'we quickly walked' to 'we walked', but also using delay will 
similarlynessly match it to 'wequickly wlkd', I will check if a node is missed 
1 step for missing the word quickly, AND will check in a node if its parts are 
missing i.e. the wlkd is missing someparts.

Thanks for enlightening me on this matter, I had not realized this.
------------------------------------------
Artificial General Intelligence List: AGI
Permalink: 
https://agi.topicbox.com/groups/agi/T90b7756a48658254-Me5371d54e10e89f90706c6ed
Delivery options: https://agi.topicbox.com/groups/agi/subscription

Reply via email to