Triple
T28721944
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Eddie Kaspbrak (It, 2017 film) |
E730118
|
entity |
| Predicate | childVersionPortrayedBy |
P39940
|
FINISHED |
| Object | Jack Dylan Grazer |
—
|
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: Jack Dylan Grazer | Statement: [Eddie Kaspbrak (It, 2017 film), childVersionPortrayedBy, Jack Dylan Grazer]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: childVersionPortrayedBy Context triple: [Eddie Kaspbrak (It, 2017 film), childVersionPortrayedBy, Jack Dylan Grazer]
-
A.
portraysYoungerVersionOfCharacterFrom
Indicates that one character is depicted as a younger version of another character from a specified source.
-
B.
childRepresents
Indicates that a child entity serves as a representation, proxy, or stand-in for another entity in some context or structure.
-
C.
depictsChild
Indicates that one entity visually represents or portrays another entity in the role or state of being a child.
-
D.
youngerVersionPortrayedBy
chosen
Indicates that one person portrays a younger version of another person, typically in a film, television show, or similar narrative work.
-
E.
childIs
Indicates that one entity is the child (offspring or subordinate descendant) of 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_69f043e91fe48190b73bcd8e08d433e0 |
completed | April 28, 2026, 5:21 a.m. |
| NER | Named-entity recognition | batch_69f658ee40088190b71e1219407690d0 |
completed | May 2, 2026, 8:05 p.m. |
| PD | Predicate disambiguation | batch_69f65760fd3081908ffe014a5e2bf069 |
completed | May 2, 2026, 7:58 p.m. |
Created at: April 28, 2026, 5:54 a.m.