Triple
T23283924
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Total Rewards |
E588937
|
entity |
| Predicate | higherTierProvides |
P151665
|
FINISHED |
| Object | increased earning rates |
—
|
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: increased earning rates | Statement: [Total Rewards, higherTierProvides, increased earning rates]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: higherTierProvides Context triple: [Total Rewards, higherTierProvides, increased earning rates]
-
A.
hasUpperTier
Indicates that one entity occupies a higher or superior level, rank, or layer in a hierarchical structure relative to another entity.
-
B.
higherTierAuthority
Indicates that one entity holds a superior or higher-ranking level of authority or control over another entity.
-
C.
hasLowerTier
Indicates that one entity is ranked, valued, or classified at a lower level or tier relative to another entity.
-
D.
topTierQualifiesFor
Indicates that entities in the highest or top tier meet the necessary criteria to be eligible for a specified benefit, status, or action.
-
E.
topTierUntil
Indicates that an entity holds a top-tier or highest-status position only up to a specified point in time, after which this status no longer applies.
- 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_69e25d16e2c08190a291de254703129e |
completed | April 17, 2026, 4:17 p.m. |
| NER | Named-entity recognition | batch_69f196454b4c8190a797537ce8912241 |
completed | April 29, 2026, 5:25 a.m. |
| PD | Predicate disambiguation | batch_69effcecabd88190856fb6e1d993e4dd |
completed | April 28, 2026, 12:18 a.m. |
| PDg | Predicate description generation | batch_69f01d8770d081908897c28b04e5faea |
completed | April 28, 2026, 2:37 a.m. |
Created at: April 17, 2026, 4:58 p.m.