Do You Think This Paradime Should Work for More Structured Complex S Like Unit Tests?

If you have a set of expected behaviours, you should be able to write them down. Even if it's harder to write perturbations, sanity checks should abound for every one of these. It i specifically sanity checks test ind specific things like names and negation and what not. If your system deals with language, it should deal with these capabilities in language. Do you think this paradime works for more structured complex s like, for instance, dependency pursing? I think it should work for anything in an l p where you can specify behavior. So basically anything. We've talked a lot about s sentiment analysis, and a little bit about reading comprehension style stuff. Be case.