Triple
T25475084
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Grail |
E638409
|
entity |
| Predicate | countryOfOriginOfSystem |
P26
|
FINISHED |
| Object | Soviet Union |
—
|
NE NERFINISHED |
How this triple was built (2 steps)
Every LLM step that produced this triple, in pipeline order — named-entity classification, the disambiguation choices (the exact options shown, with the pick highlighted), and the generated description. The batch + timestamp of each is in the Provenance table below.
NER
Named-entity recognition
gpt-5-mini
Instruction
Given a phrase, classify it is english named entity (e.g., persons, organizations, works of art) in Latin script, or not (e.g., literals, dates, URLs, verbose phrases). For disambiguation, the statement where the phrase occurs as object is also given. Please return a JSON object with `phrase` (string, the phrase being analyzed) and `is_ne` (boolean, indicating whether the phrase is a Named Entity).
Input
Phrase: Soviet Union | Statement: [Grail, countryOfOriginOfSystem, Soviet Union]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: countryOfOriginOfSystem Context triple: [Grail, countryOfOriginOfSystem, Soviet Union]
-
A.
countryOfOrigin
chosen
Indicates the country from which an entity originally comes or was first produced, created, or established.
-
B.
engineOriginCountry
Indicates the country where an engine was originally designed, manufactured, or first produced.
-
C.
organizationCountryOfOrigin
Indicates the country where an organization was originally founded or established.
-
D.
placeOfOrigin
Indicates the location or source from which an entity originally comes or was created.
-
E.
originallyProducedIn
Indicates that something was first created, manufactured, or brought into existence in a particular place or location.
- F. None of above.
Provenance (3 batches)
The batch behind each pipeline step, in order, with when it ran. Timestamps are batch-level — stages were processed in waves, so the object chain (NER → NED1 → NEDg → NED2) reads in order, but predicate / elicitation batches can sit in a different wave.
| Step | Stage | Batch ID | Status | When |
|---|---|---|---|---|
| creating | Elicitation | batch_69e75db9b964819096802dcf502e577e |
completed | April 21, 2026, 11:21 a.m. |
| NER | Named-entity recognition | batch_69f68805b4848190b75da14996d52a38 |
completed | May 2, 2026, 11:25 p.m. |
| PD | Predicate disambiguation | batch_69f68609c0b08190a8e1238a4d97c270 |
completed | May 2, 2026, 11:17 p.m. |
Created at: April 21, 2026, 2:25 p.m.