it makes sense for the buffer to be searchable. So at any given time only some information is actually provided as an input parameter to the model, but as it “thinks” serially it can order a search.
For example, a task like “write a large computer program”. It cannot remember all the variable names and interfaces to the other parts of the program it is not working on, but needs to locate them whenever it calls on them.
The obvious requirement is for the AI to have a second buffer it can write to as ‘scratch’ to keep track of the information for tasks like this.
So long as the “buffer” is a set of parameters/weights/neurons, that would fit my test.
it makes sense for the buffer to be searchable. So at any given time only some information is actually provided as an input parameter to the model, but as it “thinks” serially it can order a search.
For example, a task like “write a large computer program”. It cannot remember all the variable names and interfaces to the other parts of the program it is not working on, but needs to locate them whenever it calls on them.