Triple
T265
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Carnegie Institution of Washington |
E4
|
entity |
| Predicate | hasChairperson |
P10
|
FINISHED |
| Object | board of trustees (Carnegie Institution of Washington) |
—
|
LITERAL FINISHED |
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: board of trustees (Carnegie Institution of Washington) | Statement: [Carnegie Institution of Washington, hasChairperson, board of trustees (Carnegie Institution of Washington)]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: hasChairperson Context triple: [Carnegie Institution of Washington, hasChairperson, board of trustees (Carnegie Institution of Washington)]
-
A.
positionHeld
Indicates that an entity occupies or has occupied a specific role, job, office, or position within an organization or context.
-
B.
memberOf
chosen
Indicates that an entity belongs to, is part of, or is a constituent of a larger group, organization, or collection.
-
C.
notableFor
Indicates that an entity is especially recognized or distinguished for a particular quality, achievement, characteristic, or role.
-
D.
givenName
Indicates the personal first name assigned to an individual.
-
E.
fullName
Indicates that an entity has a complete personal name, typically combining given name(s) and family name into a single string.
- 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_69a222a954e48190b48f126a67485661 |
completed | Feb. 27, 2026, 11:03 p.m. |
| NER | Named-entity recognition | batch_69a2266edf048190828e8f53cb7f6ba6 |
completed | Feb. 27, 2026, 11:19 p.m. |
| PD | Predicate disambiguation | batch_69a222f9916081908db2eedc81d85301 |
completed | Feb. 27, 2026, 11:04 p.m. |
Created at: Feb. 27, 2026, 11:04 p.m.