Triple
T1890966
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | 2010 Miracle at the New Meadowlands comeback |
E41871
|
entity |
| Predicate | deficitOvercomePoints |
P11634
|
FINISHED |
| Object | 21 |
—
|
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: 21 | Statement: [2010 Miracle at the New Meadowlands comeback, deficitOvercomePoints, 21]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: deficitOvercomePoints Context triple: [2010 Miracle at the New Meadowlands comeback, deficitOvercomePoints, 21]
-
A.
seriesDeficitOvercome
Indicates that a competitor or team has successfully come back from a losing position in a multi-game series to erase the deficit.
-
B.
largestDeficitOvercome
chosen
Indicates the maximum deficit or point difference that was successfully overcome by one side to achieve a comeback in a contest or competition.
-
C.
winnerPoints
Indicates the number of points earned by the winning participant or entity in a competition or event.
-
D.
loserPoints
Indicates the number of points awarded to or accumulated by the losing side in a competitive event or comparison.
-
E.
overcameHoldingOf
Indicates that one entity successfully defeated, reversed, or nullified another entity’s control, possession, or dominance over something.
- 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_69a8864b6de0819098d089f6a1b910a7 |
completed | March 4, 2026, 7:21 p.m. |
| NER | Named-entity recognition | batch_69abb14475448190b291ada3454bf98b |
completed | March 7, 2026, 5:01 a.m. |
| PD | Predicate disambiguation | batch_69abafe61bc48190ac9ead027df930e1 |
completed | March 7, 2026, 4:56 a.m. |
Created at: March 4, 2026, 7:34 p.m.