Glossary
The words that come up when you read about System One AI models, both decision models and extraction models, defined once and linked from everywhere else on the site.
B
C
- Calibrated ProbabilityA calibrated probability is a number from a model that means what it says: across all the answers given at 80%, about 80% are right.
- CalibrationCalibration is the match between a model's stated probabilities and how often it turns out to be right.
- ChoiceA Choice is a question type in decision models such as Jev, for selecting one option from a defined set.
- Classification modelA classification model assigns an input to one of a fixed set of labels, and usually returns a probability for each label.
- Composite scoringComposite scoring is one of TypeSafe's documented patterns, described as "Combine several dimensions of analysis into a single score." You ask a separate question about each dimension, then your own code weights and combines the answers.
- ConfidenceConfidence is a number from 0 to 1 that TypeSafe computes from the probability distribution an answer already carries.
- Confidence GatingConfidence gating is the practice of branching your code on a model's confidence value as well as on its answer.
D
E
- EstimateAn estimate is a Clairic field type that predicts a value the input does not settle, such as when an invoice will be paid.
- Expected calibration errorExpected calibration error (ECE) is a score for how far a model's confidence is from its real hit rate: a model that says 80% should be right about 80% of the time.
- Extraction modelAn extraction model reads text or images and returns typed values instead of decisions or prose: a span of the input, a number, a date, a duration, a list, an estimate or a forecast, each with probabilities.
F
H
I
J
N
- Non-autoregressiveNon-autoregressive describes a model that does not build its output one token at a time.
- NoulNoul meaning: a Noul is a yes/no question type in System One decision models such as Jev.
- Null thresholdA null threshold is the p_null above which an extraction model returns null for a field instead of a value.
P
Q
R
S
- ScoreA Score is a question type in decision models such as Jev that rates content against ordered, descriptive levels, minimum 2 and maximum 10.
- SpanA span is a stretch of the input text, identified by its start and end offsets.
- Speculative fan-outSpeculative fan-out is one of TypeSafe's documented patterns: "Send many questions in a single call, including speculative ones, and let your code decide what's relevant." Because questions are evaluated in parallel, extra questions typically cost no extra latency, so you ask ahead instead of making a second round trip.
- StateState is TypeSafe's word for the content you hand a System One model: the ticket, document, transcript or JSON object the questions are asked about.
- Structured outputsStructured outputs is the existing technique for making a language model emit valid JSON: constrain decoding to a schema so every generated token keeps the output parseable.
- System One ModelA System One model is an AI model that returns typed answers with probabilities in one fast pass, instead of generated text.
- System Two ModelSystem Two model is a contrast term, not a product.
T
- Thinking, Fast and SlowThinking, Fast and Slow is Daniel Kahneman's 2011 book (Farrar, Straus and Giroux) describing two modes of human thought: a fast, automatic System 1 and a slow, deliberate System 2.
- Typed outputTyped output is a model answer that arrives already shaped as a value your code can use, such as an enum member or a probability between 0 and 1.
- TypeSafe AITypeSafe AI is the San Francisco company that coined the term System One model and shipped the first one, Jev, on 2026-09-15.
Z
- Zero-shot classification modelA zero-shot classification model sorts text into labels you supply at call time, without being trained on those labels.
- Zero-shot classifierA zero-shot classifier assigns text to labels it was never trained on, using the label descriptions you supply at call time rather than labelled examples.