Triple
T8583453
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Arrow Cross Party |
E203241
|
entity |
| Predicate | responsibleForDeathOf |
P74108
|
FINISHED |
| Object | tens of thousands of Hungarian Jews |
—
|
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: tens of thousands of Hungarian Jews | Statement: [Arrow Cross Party, responsibleForDeathOf, tens of thousands of Hungarian Jews]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: responsibleForDeathOf Context triple: [Arrow Cross Party, responsibleForDeathOf, tens of thousands of Hungarian Jews]
-
A.
allegedToHaveKilled
Indicates that one entity is claimed or accused, but not proven, to have killed another entity.
-
B.
perpetratorOfKilling
chosen
Indicates that an entity is the one who carried out or caused a particular killing.
-
C.
assassinatedIn
Indicates that an assassination of one entity occurred within the specified location or context represented by another entity.
-
D.
killedBy
Indicates that one entity caused the death of another entity.
-
E.
murderVictimOf
Indicates that one entity is the person who was killed by another entity in an act of murder.
- 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_69ca8329bb7c8190a63c643730839103 |
completed | March 30, 2026, 2:05 p.m. |
| NER | Named-entity recognition | batch_69cbeb1d0edc8190b4495935275252f3 |
completed | March 31, 2026, 3:41 p.m. |
| PD | Predicate disambiguation | batch_69cbd11b13108190b07f8f161425a585 |
completed | March 31, 2026, 1:50 p.m. |
Created at: March 30, 2026, 6:22 p.m.