Is there a tutorial or other source that describes why, when, and how to use more than one vocab?
For the new Nengo SPA module, there is some documentation on this topic. Though, it doesn’t say too much about your specific question.
In my counting model, I used two separate vocabs. One for controlling the information flow through the model (containing RUN and NONE) and one for the numerical representation (containing digitis ZERO to NINE). Basically, it is typical to want different vocabs when you have different domains of computation.
@xchoo do you know of a good example of this in Spaun?