Triple
T114129
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Zyklon B |
E2306
|
entity |
| Predicate | mechanismOfAction |
P1634
|
FINISHED |
| Object | inhibition of cellular respiration |
—
|
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: inhibition of cellular respiration | Statement: [Zyklon B, mechanismOfAction, inhibition of cellular respiration]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: mechanismOfAction Context triple: [Zyklon B, mechanismOfAction, inhibition of cellular respiration]
-
A.
removalMechanism
Indicates the method or process by which something is removed, eliminated, or taken away from a system or context.
-
B.
transitionMechanismUsedWith
Indicates that a particular transition or change from one state to another is carried out using a specified mechanism or method.
-
C.
primaryEffect
chosen
Indicates the main direct outcome or consequence that results from a given cause, action, or condition.
-
D.
remedy
Indicates that one entity serves to cure, alleviate, or counteract a problem, illness, or undesirable condition affecting another entity.
-
E.
isMechanicalOrElectronic
Indicates that something operates using mechanical components, electronic components, or a combination of both.
- 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_69a24fcdaeb48190a2d796677e4b3281 |
completed | Feb. 28, 2026, 2:15 a.m. |
| NER | Named-entity recognition | batch_69a25760af348190bf402089c240887d |
completed | Feb. 28, 2026, 2:48 a.m. |
| PD | Predicate disambiguation | batch_69a2564417848190a8a8a38e97348963 |
completed | Feb. 28, 2026, 2:43 a.m. |
Created at: Feb. 28, 2026, 2:20 a.m.