Compositionality

Getting the right contents.

• semantic capacity tested — knowing the propositional content
• operationalization — form => meaning mapping
• measure of success — generalizing to the right meaning representation for novel expressions (this is non-trivial; multiple compatible generalization maybe applicable depending on context)

task

Given that a model can map certain expressions to their meaning representations, can they also do this for new expressions?

results

• lexical generalization (fill in new words/pairs) isn’t too hard for new NN

Its like a transforming distributions procedure, but your \(f\) is not constrained to be differentiable. So you can still sample from it.

we perform a random sample of possible next state (weighted by the action you took, meaning an instantiation of \(s’ \sim T(\cdot | s,a)\)) and reward \(R(s,a)\) from current state

a hissyfight with the transformational generative syntax.

generative semantics states that structure is in support of meaning, rather than the other way around that transformational generative syntax suggests.

This means that you need to first come up with a meaning then imbew the best structure to support the expression of that meaning.

This (along with distributed morphology) is the main opposition of the Lexicalist Hypothesis, and because proof for the existence of semantic primes, also the main opposition of the existence of semantic primes.