Triple
T30679965
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Earth’s magnetic field |
E781025
|
entity |
| Predicate | deflects |
P85627
|
FINISHED |
| Object | charged particles from the Sun |
—
|
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: charged particles from the Sun | Statement: [Earth’s magnetic field, deflects, charged particles from the Sun]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: deflects Context triple: [Earth’s magnetic field, deflects, charged particles from the Sun]
-
A.
defends
Indicates that one entity protects or supports another entity against attack, criticism, or harm.
-
B.
deflectionType
chosen
Indicates the specific manner or category of how something is diverted, redirected, or caused to change its original course or direction.
-
C.
reflects
Indicates that one entity (often a surface, medium, or representation) throws back, mirrors, or otherwise shows an image, property, or state of another entity.
-
D.
reflector
Indicates that one entity functions as a surface or device that reflects light, sound, or other waves originating from another entity.
-
E.
defender
Indicates a relationship where one entity protects, guards, or supports another entity against threats, attacks, or criticism.
- 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_69f224a92f54819095499b4d32bd5134 |
completed | April 29, 2026, 3:32 p.m. |
| NER | Named-entity recognition | batch_69f68b1b21308190a63dc200654683fc |
completed | May 2, 2026, 11:39 p.m. |
| PD | Predicate disambiguation | batch_69f6861170d08190bb98be609d436f84 |
completed | May 2, 2026, 11:17 p.m. |
Created at: April 29, 2026, 8:32 p.m.