Triple
T25182199
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Atlas Passage |
E630611
|
entity |
| Predicate | locatedOnOrOff |
P40
|
FINISHED |
| Object | Istiklal Avenue |
—
|
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: Istiklal Avenue | Statement: [Atlas Passage, locatedOnOrOff, Istiklal Avenue]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: locatedOnOrOff Context triple: [Atlas Passage, locatedOnOrOff, Istiklal Avenue]
-
A.
locatedOff
Indicates that one entity is situated away from, but in the vicinity of, another entity or reference point, rather than directly on or within it.
-
B.
locatedIn
chosen
Indicates that one entity exists or is situated within the spatial, administrative, or conceptual boundaries of another entity.
-
C.
locatedOnFront
Indicates that one entity is positioned on the front side or face of another entity.
-
D.
isPositionOn
Indicates that one entity occupies or corresponds to a specific location or coordinate relative to another reference frame or space.
-
E.
isLocatedToThe
Indicates a spatial relationship where one entity is positioned in a specific directional orientation relative to another entity.
- 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_69e75a88fdf081908e47ae6e195c14e1 |
completed | April 21, 2026, 11:07 a.m. |
| NER | Named-entity recognition | batch_69f6aaf50be08190a2b62a6d881f8aee |
completed | May 3, 2026, 1:55 a.m. |
| PD | Predicate disambiguation | batch_69f6aa1c555081908787dbf76147f180 |
completed | May 3, 2026, 1:51 a.m. |
Created at: April 21, 2026, 12:36 p.m.