/ 01
Short Answer
Noul is one of Jev’s three question types. You ask a yes/no question, and Jev returns a single number from 0 to 1: the probability that the answer is yes. Zero means no and one means yes. Your application chooses the threshold for acting on that probability.
/ 02
Noul, Choice and Score — The Three Question Types
| Type | You provide | Jev returns | Use for |
|---|---|---|---|
| Noul | Yes/no question, optional true/false criteria | One probability, 0–1 | Checks and guardrails |
| Choice | Named options and descriptions, up to 255 | Selected option, distribution, confidence | Classification and routing |
| Score | 2–10 ordered level descriptions | Weighted score, legend, distribution, confidence | Ratings and severity |
/ 03
The Noul Request
{
"state": "I have asked three times. Can I talk to a person?",
"model": "jev-latest",
"questions": {
"is_human_escalation": { "type": "noul", "instructions": "Is the customer asking for a human agent?" },
"is_repeat_contact": {
"type": "noul", "instructions": "Has the customer contacted support before?",
"criteria": { "true": "Mentions a prior attempt or ticket", "false": "No sign of previous contact" }
}
}
}You name the question ids. TypeSafe says those ids are not used in inference; each answer returns under the same id.
/ 04
The Noul Answer
{ "type": "noul", "noul": 0.95 }This is a 95% probability of yes. Noul answers carry no confidence value and no probability distribution — just the single number.
/ 05
When to Use Criteria
Without criteria, Jev uses the instructions alone. Add true and false descriptions when a boundary is ambiguous, such as what counts as repeat contact. Keep descriptions aligned with the question.
/ 06
Choosing Your Threshold
Jev gives a probability; your code selects the action. A 0.90, 0.95 or 0.99 threshold trades automation coverage for stricter evidence. A practical three-band policy can approve high values, review the middle and reject low values. Validate each band on your data, especially for high-risk actions.
/ 07
When Noul Is the Wrong Tool
Use Choice for one of several named answers and Score for a degree on an ordered scale. Jev 1.13 also has documented weaknesses in counting, dates, indirection and structural assumptions; do not hide calculations in a Noul. Read the known limitations and official jaggedness notes.
/ 08
Noul vs Asking an LLM
A text-generating model may answer in prose or outside a requested set. Noul has one bounded output from 0 to 1, making threshold logic explicit. It is still a model judgment that needs validation on your own data.
/ 09
Try Noul in the Playground
Open the Jev playground and choose Noul. You can ask multiple atomic questions per call, which TypeSafe recommends. See Jev as a classifier for a support queue example.
/ FAQ
Frequently asked questions
What does “noul” mean in Jev?
It is Jev’s yes/no question primitive.
Is a noul answer a boolean?
No. It is a probability from 0 to 1.
Does a noul answer include a confidence score?
No. Noul returns only a single probability.
What noul value should I treat as true?
Choose and validate a threshold for your use case; higher-risk actions usually need stricter cutoffs.
Can I send several noul questions in one request?
Yes. The questions map can contain several ids and a mix of all three types.
Is the word noul used anywhere else?
It can appear in other contexts, including medical terminology. Here it specifically means the Jev primitive.