Triple
T206588
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Atlantic Meridional Overturning Circulation |
E4621
|
entity |
| Predicate | canExhibit |
P9982
|
FINISHED |
| Object | abrupt changes |
—
|
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: abrupt changes | Statement: [Atlantic Meridional Overturning Circulation, canExhibit, abrupt changes]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: canExhibit Context triple: [Atlantic Meridional Overturning Circulation, canExhibit, abrupt changes]
-
A.
hasExhibits
Indicates that an entity (such as a museum, gallery, or event) displays or presents certain items, artworks, or objects as part of its collection or show.
-
B.
hasExhibition
Indicates that an entity organizes, hosts, or presents a particular exhibition.
-
C.
hasInteractiveExhibits
Indicates that something contains exhibits designed for active participation or engagement by the audience.
-
D.
numberOfExhibits
Indicates the total count of exhibits associated with a given entity or context.
-
E.
exhibitsWorkOf
Indicates that one entity displays or presents the creative works produced by another entity.
- F. None of above. chosen
Provenance (4 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_69a25737567c81908f9c505300239181 |
completed | Feb. 28, 2026, 2:47 a.m. |
| NER | Named-entity recognition | batch_69a25e2aba74819093eddd8d820260c0 |
completed | Feb. 28, 2026, 3:16 a.m. |
| PD | Predicate disambiguation | batch_69a25b4c7f908190876c1041db52dffc |
completed | Feb. 28, 2026, 3:04 a.m. |
| PDg | Predicate description generation | batch_69a25e292fdc8190bfd51d8848f9ed58 |
completed | Feb. 28, 2026, 3:16 a.m. |
Created at: Feb. 28, 2026, 2:51 a.m.