Share
Glossary/Evidence grades

Training

Background the model already knew rather than something it read during this run. It is kept because it is usually load-bearing context, and it is labelled because nothing about the run confirms it. Where the schema requires one, a ready-made search rides along so you can check it in one tap.

The definition the product enforces, read from the same code the methodology page renders.

How to read it

When Training appears on a page

Background the model already held rather than something it fetched this run. It is kept because it is usually the context the rest stands on, and labelled so you know nothing in the run confirmed it. Where the schema requires it, a ready-made search sits beside the row.

The family

Evidence grades

The grade every claim on a research page carries: how it is known, not how confident it sounds. Assigned by the same rules on every scan; the rollup numbers are computed from the graded rows.

VerifiedCorroborated against an outside source, and that source is linked.GroundedCame from material the scan actually read, with a named source.PartialSome support, not enough to call it settled.ContestedSources disagree, or the record contradicts it.UnverifiedAsserted in the material, with nothing behind it.Your inputYou supplied it. The scan took it as given.From the recordPresent in the material with no provenance attached.GapSomething the analysis could not establish, reported as missing.
Where you will see it

On these pages

Kinds of research that print it
Event →Inquiry →Open Case →Scholar →Market Research →
Surfaces
Every research page's FACTS tab and footer
The Sources sidebar
The corroboration view of your own research
The rubric

The whole grading vocabulary, on one page

Every grade here is assigned by the same rules on every scan, and the methodology page renders those rules from the code that enforces them.

How the grades work →The whole glossary