Triple
T15600945
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Ásbrú |
E375029
|
entity |
| Predicate | associatedWithSound |
P22084
|
FINISHED |
| Object | Heimdallr’s horn Gjallarhorn at its defense |
—
|
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: Heimdallr’s horn Gjallarhorn at its defense | Statement: [Ásbrú, associatedWithSound, Heimdallr’s horn Gjallarhorn at its defense]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: associatedWithSound Context triple: [Ásbrú, associatedWithSound, Heimdallr’s horn Gjallarhorn at its defense]
-
A.
followsSoundOf
Indicates that one entity moves or directs its attention by tracking the source or direction of a sound produced by another entity.
-
B.
hasSound
chosen
Indicates that an entity produces, emits, or is associated with a particular sound.
-
C.
associatedWithGround
Indicates that an entity has a defined connection or relationship with the ground, such as contact, support, or reference.
-
D.
includesAddedSound
Indicates that one entity contains or incorporates an additional sound element beyond the original or primary audio content.
-
E.
associatedWithTrack
Indicates a relationship in which an entity is linked or connected to a specific track, such as being part of, derived from, or relevant to that track.
- 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_69d85cce25008190b13b52745fbd719b |
completed | April 10, 2026, 2:13 a.m. |
| NER | Named-entity recognition | batch_69e04e621fc4819097e8e85e7ddfdc6c |
completed | April 16, 2026, 2:50 a.m. |
| PD | Predicate disambiguation | batch_69deda844af081909e658ebc9d9b403d |
completed | April 15, 2026, 12:23 a.m. |
Created at: April 10, 2026, 4:12 a.m.