So I read the title and was like “huh, that’s a novel and interesting idea”. And then I read the description and was like “dude, that is not what ‘compositional’ means”.
Fair point, I’m using “compositional” in an informal sense different from the one in formal semantics, closer to what I called “trivial compositionally” in this paper. But I’d argue it’s not totally crazy to call such preference models compositional and that compositionally here still has some resemblance to Montague’s account of compositionally as homeomorphism: basically, you have get_total_score(response) == sum([get_score(attribute) for attribute in decompose(response)])
So I read the title and was like “huh, that’s a novel and interesting idea”. And then I read the description and was like “dude, that is not what ‘compositional’ means”.
Fair point, I’m using “compositional” in an informal sense different from the one in formal semantics, closer to what I called “trivial compositionally” in this paper. But I’d argue it’s not totally crazy to call such preference models compositional and that compositionally here still has some resemblance to Montague’s account of compositionally as homeomorphism: basically, you have
get_total_score(response) == sum([get_score(attribute) for attribute in decompose(response)])