Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →Choose Jev’s primitive by the shape of the answer your application needs: use Choice for one label from unordered alternatives, Score for a position on a defined ordered scale, and Noul for the probability that one yes-or-no statement is true. The distinctions matter because each primitive returns a different kind of result and supports different downstream decisions.
TypeSafe AI’s official Primitives (Questions) documentation describes the three as distinct typed questions whose answers you can combine in application code.
As an Amazon Associate I earn from qualifying purchases.
How to choose among Choice, Score and Noul
Start with what the application will do with the answer—not with which primitive sounds most familiar.
| Primitive | Answer shape | Use it when | Uncertainty signal |
|---|---|---|---|
| Choice | A selected option, probabilities across the options and confidence | The answer is one label from an unordered set, such as a category used to route a request. | Confidence and the option probability distribution. |
| Score | A position on defined ordered levels, with level probabilities and confidence | The answer belongs on a meaningful spectrum or rubric, such as severity. | Confidence and the level probability distribution. Treat the result as a probability-weighted assessment, not an exact measurement. |
| Noul | The probability that one statement is true | You need a yes-or-no signal about one proposition. | A probability near 0.5 means true and false are about equally likely; Noul has no separate confidence field. |
For example, “Which department should handle this request?” suggests Choice if the departments are simply alternatives. “How severe is the issue?” suggests Score if you define ordered levels. “Is the report urgent?” suggests Noul if the answer is whether that single statement is true.
Mistake 1: Using Choice for an ordered scale
Choice is for distinct alternatives, not for levels whose order carries meaning. If the options are “low,” “medium” and “high,” the answer is a position on a scale. Define those levels as ordered Score levels so the application can interpret their relationship rather than treating them as unrelated labels.
Mistake 2: Using Score for unrelated labels
Use Score only when levels have a meaningful order. If the possible answers are separate categories with no meaningful in-between—such as “billing,” “account access” or “technical support”—use Choice. A score implies an ordered relationship; it is a poor fit when a category is simply one member of a set.
Define the categories
For Choice, provide the option map your application expects. The official Choice guide says a question can accept up to 255 options and recommends including “other” or “none of the above” when the set may be incomplete. That gives the model a defined way to represent an answer outside the usual categories.
Mistake 3: Reading Noul as a measure of degree
Noul answers whether one proposition is true; it does not measure the intensity, quality or urgency of a thing. A result of 0.5 is not “medium urgency” or “average quality.” TypeSafe AI’s official Primitives (Questions) documentation puts it this way: “A Noul value of 0.5 means the model gives yes and no equal probability.” Use Score with described ordered levels when the application needs a degree or rubric position.
Rank #3
Mistake 4: Combining multiple judgments in one question
A Noul question should express one proposition. If you ask whether a message is both urgent and about billing, a single yes-or-no probability does not make clear which condition drove the result. Ask separate questions—one about urgency, one about the billing category—and handle each answer according to its type.
Combine answers in application code
TypeSafe’s documentation explains that questions sharing the same state can be evaluated independently in a request and their answers combined in application code. Do not assume that one question’s answer silently becomes context for another. Make the relationship explicit in your code—for example, route a message only when the urgency result and category result each meet the criteria your application defines.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Match the answer to the downstream action
- Route or branch on a category: use Choice when the destination is one label from an unordered set.
- Compare against a threshold: use Score when the result represents a position on a described ordered scale.
- Act on a yes-or-no signal: use Noul for the probability that one proposition is true.
That mapping keeps the result interpretable: a label remains a label, an ordered assessment remains a score, and a binary proposition remains a probability rather than being mistaken for a measure of degree.
Recommended Free Tools
Quick Recap
Best Value
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
