Testing compositionality on GCCKR

can we just represent each prime as a point in a vector?

We would need to have a programming language that could interpret the vector. So we are moving the meaning of the vector to the interpreter.

We need to test compositionality. Can you have multiple vectors as a sequence to represent new concepts/words? The point of new words is to act like a cognitive short cut, we get to new/different meaning faster. So I’m not sure if combining words/concepts will do anything by itself. The architecture must allow for some sort of advantage by connecting new concepts.

The interpreter could do one of several things.

Generate scenes based off those words.

generate definitions of those words.

Could we ask questions to test if its working?

So its a vector representation + a visual generator.

Leave a comment

Your email address will not be published. Required fields are marked *