Triple
T29858860
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | NVLink interconnect |
E758260
|
entity |
| Predicate | bandwidthAdvantageOver |
P167674
|
FINISHED |
| Object | PCIe 3.0 |
—
|
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: PCIe 3.0 | Statement: [NVLink interconnect, bandwidthAdvantageOver, PCIe 3.0]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: bandwidthAdvantageOver Context triple: [NVLink interconnect, bandwidthAdvantageOver, PCIe 3.0]
-
A.
bandwidthCharacteristic
Indicates the relationship specifying the bandwidth-related properties or constraints associated with a connection, link, or communication channel.
-
B.
bandwidthEfficiency
Indicates how effectively available bandwidth is utilized to transmit data relative to capacity or resource usage.
-
C.
increasesBandwidthToward
Indicates that one entity causes an increase in the available data transmission capacity directed toward another entity.
-
D.
bandwidthClass
Indicates the classification of a connection or resource based on its available or allocated bandwidth capacity.
-
E.
supportsBandwidths
Indicates that an entity is compatible with or can operate using the specified range or set of bandwidth values.
- 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_69f2245b4dec8190b85f664d918a00a5 |
completed | April 29, 2026, 3:31 p.m. |
| NER | Named-entity recognition | batch_69f67683fde88190bf2f338ec18dcaca |
completed | May 2, 2026, 10:11 p.m. |
| PD | Predicate disambiguation | batch_69f66ac32b60819092290b2de35988d3 |
completed | May 2, 2026, 9:21 p.m. |
| PDg | Predicate description generation | batch_69f66bd123108190b451eb6e23842adb |
completed | May 2, 2026, 9:25 p.m. |
Created at: April 29, 2026, 5:48 p.m.