Triple
T1793180
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Atari deep Q-network |
E39543
|
entity |
| Predicate | usesStabilizationTechnique |
P3047
|
FINISHED |
| Object | experience replay |
—
|
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: experience replay | Statement: [Atari deep Q-network, usesStabilizationTechnique, experience replay]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: usesStabilizationTechnique Context triple: [Atari deep Q-network, usesStabilizationTechnique, experience replay]
-
A.
hasOpticalImageStabilization
Indicates that a device or component includes a feature that reduces image blur caused by camera movement during capture.
-
B.
hasTechnique
chosen
Indicates that an entity employs, utilizes, or is associated with a particular method, procedure, or technique.
-
C.
stabilizedBy
Indicates that an entity’s state, structure, or behavior is made more steady, secure, or resistant to change through the influence or support of another entity.
-
D.
isStable
Indicates that the state, condition, or configuration of an entity does not change significantly over time or under expected variations in its environment.
-
E.
gimbalCapability
Indicates the ability of a system or device to support and control a gimbal’s movement or stabilization functions.
- 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_69a88631854081909723959921e45c2b |
completed | March 4, 2026, 7:21 p.m. |
| NER | Named-entity recognition | batch_69ab61b6ea188190aab9fb839bf1e367 |
completed | March 6, 2026, 11:22 p.m. |
| PD | Predicate disambiguation | batch_69aa61d2f7a8819090301f92d3e358c7 |
completed | March 6, 2026, 5:10 a.m. |
Created at: March 4, 2026, 7:32 p.m.