Triple
T10423877
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Arctic fox |
E245732
|
entity |
| Predicate | scavenges |
P93511
|
FINISHED |
| Object | polar bear kills |
—
|
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: polar bear kills | Statement: [Arctic fox, scavenges, polar bear kills]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: scavenges Context triple: [Arctic fox, scavenges, polar bear kills]
-
A.
hauls
Indicates that one entity pulls, drags, or transports another entity, typically involving effort or force over a distance.
-
B.
scrapped
Indicates that something planned, created, or in progress was abandoned, discarded, or cancelled and will no longer be pursued or used.
-
C.
scans
Indicates performing a systematic examination or inspection of something, often using a device or method to detect, read, or analyze its contents or structure.
-
D.
largelyConsumedBy
Indicates that something is mostly or predominantly eaten or used up by a particular consumer or group.
-
E.
catches
Indicates that one entity successfully seizes, intercepts, or takes hold of another entity, often stopping its motion or preventing its escape.
- 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_69d381bf3dc08190bf35a2643e4e8f22 |
completed | April 6, 2026, 9:49 a.m. |
| NER | Named-entity recognition | batch_69d4ea2de4d48190aee65b3f6ec3cc48 |
completed | April 7, 2026, 11:27 a.m. |
| PD | Predicate disambiguation | batch_69d4dfb9d3648190aaabed901f22a8c0 |
completed | April 7, 2026, 10:43 a.m. |
| PDg | Predicate description generation | batch_69d4e91de3088190ae89cbb12fb29db8 |
completed | April 7, 2026, 11:23 a.m. |
Created at: April 6, 2026, 12:12 p.m.