llm

according to media 
a llm is a bunch of parameters in a layer 
where each parameter connects to parameters in another layer 
and there are many parameters in each layer 
and there are many layers in a llm 
but google search does not show web articles 
that explain how many connections each parameter in a llm has

and a human brain is neurons connecting to each other 
but probably not in layers 
and each neuron connects to a lot of other neurons 

so if a llm is made more brain like 
then maybe it could become more intelligent 

maybe by each parameter having more connections to other parameters 
maybe by having clusters of parameters instead of layers of parameters 
maybe by allowing llm to self organize it's clusters of parameters 

example: 
one cluster with many parameters 
with few connections to each other 
that generate one functionality 

and another cluster of few parameters 
with many connections to each other 
that generate another functionality 

and a third cluster of many parameters 
with many connections to each other 
that generate yet another functionality 

example: 
in a self driving car 
one big cluster of parameters generate vision 
and a smaller cluster of parameters generate hearing 
and another cluster of parameters generate driving 

Comments

Popular posts from this blog

🖤

logic

the real big black budget