Triple
T20223943
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Marvolo Gaunt's ring |
E495326
|
entity |
| Predicate | temptation |
P139301
|
FINISHED |
| Object | ability to see dead loved ones |
—
|
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: ability to see dead loved ones | Statement: [Marvolo Gaunt's ring, temptation, ability to see dead loved ones]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: temptation Context triple: [Marvolo Gaunt's ring, temptation, ability to see dead loved ones]
-
A.
firstTemptation
Indicates the initial instance in which an entity is enticed or urged toward a potentially wrong, risky, or forbidden action.
-
B.
secondTemptation
Indicates a relationship where an entity is subjected to or engages in a second instance of temptation, following an initial tempting event.
-
C.
thirdTemptation
Indicates the relationship in which an entity is subjected to or involved in the third in a sequence of temptations or tests.
-
D.
settingOfTemptation
Indicates the location or context in which a temptation occurs or is experienced.
-
E.
involvesTempter
Indicates that the situation or event includes an entity acting as a tempter, attempting to entice or lure another into a particular action or state.
- 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_69da626cff80819097b530718a7c98b6 |
completed | April 11, 2026, 3:02 p.m. |
| NER | Named-entity recognition | batch_69e66fd827708190b798a30f4e7d533f |
completed | April 20, 2026, 6:26 p.m. |
| PD | Predicate disambiguation | batch_69e55b18609481909ab28bc8750a642f |
completed | April 19, 2026, 10:45 p.m. |
| PDg | Predicate description generation | batch_69e56702ad04819099c1c08f28d16809 |
completed | April 19, 2026, 11:36 p.m. |
Created at: April 11, 2026, 11:39 p.m.