Triple
T23698949
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Yekusiel Yehuda Halberstam |
E585532
|
entity |
| Predicate | lostFamilyIn |
P102063
|
FINISHED |
| Object | Holocaust |
—
|
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: Holocaust | Statement: [Yekusiel Yehuda Halberstam, lostFamilyIn, Holocaust]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: lostFamilyIn Context triple: [Yekusiel Yehuda Halberstam, lostFamilyIn, Holocaust]
-
A.
lostFamilyMembersIn
chosen
Indicates that a person has family members who died or went missing as a result of a specified event, place, or time period.
-
B.
lostIn
Indicates that an entity has become unable to find its way or is no longer in control or possession within a particular place, situation, or context.
-
C.
lostCompanion
Indicates that one entity has lost another entity who had been accompanying or closely associated with them.
-
D.
lostWith
Indicates that one entity experienced a loss or defeat in conjunction with, or as part of the same side/team as, another entity.
-
E.
lostAt
Indicates that an entity ceased to be in possession of or unable to be found at a specific location or context.
- 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_69e24904bd508190abfcb74855de2918 |
completed | April 17, 2026, 2:51 p.m. |
| NER | Named-entity recognition | batch_69f1b6804970819083d4e63caa93ae4e |
completed | April 29, 2026, 7:42 a.m. |
| PD | Predicate disambiguation | batch_69f155d5265881908e43a9696b6a6d0f |
completed | April 29, 2026, 12:50 a.m. |
Created at: April 17, 2026, 6:53 p.m.