Triple
T26870721
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | United States v. Bernard L. Madoff |
E676602
|
entity |
| Predicate | approximateLosses |
P180928
|
FINISHED |
| Object | tens of billions of US dollars |
—
|
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 billions of US dollars | Statement: [United States v. Bernard L. Madoff, approximateLosses, tens of billions of US dollars]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: approximateLosses Context triple: [United States v. Bernard L. Madoff, approximateLosses, tens of billions of US dollars]
-
A.
approximateEstimation
Indicates an estimation relationship where one value or assessment is only roughly or closely, but not exactly, equal to another.
-
B.
approximates
Indicates that one entity is close to, but not exactly equal to, the value, form, or behavior of another entity.
-
C.
estimatedPrincipalLoss
Indicates the amount of principal that is expected to be lost on an investment, loan, or financial position.
-
D.
casualtiesEstimate
Indicates an estimated number of people killed, injured, or otherwise harmed as a result of an event or incident.
-
E.
approximateDrop
Indicates an estimated or roughly calculated decrease in a quantity, value, or level rather than an exact measured drop.
- 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_69eee9ba94bc8190b44c5d4397d04ecd |
completed | April 27, 2026, 4:44 a.m. |
| NER | Named-entity recognition | batch_69f75dc25fa08190b371faf36d9fb72c |
completed | May 3, 2026, 2:37 p.m. |
| PD | Predicate disambiguation | batch_69f758586534819083e91172f4bf5098 |
completed | May 3, 2026, 2:14 p.m. |
| PDg | Predicate description generation | batch_69f75dc140c4819085063d6c4c36ca61 |
completed | May 3, 2026, 2:37 p.m. |
Created at: April 27, 2026, 5:32 a.m.