Triple
T34764352
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Half-Life 2: Episode One |
E1002157
|
entity |
| Predicate | AICompanion |
P181589
|
FINISHED |
| Object | Alyx Vance |
—
|
NE NERFINISHED |
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: Alyx Vance | Statement: [Half-Life 2: Episode One, AICompanion, Alyx Vance]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: AICompanion Context triple: [Half-Life 2: Episode One, AICompanion, Alyx Vance]
-
A.
navigationKnowledge
Indicates that an entity possesses information or understanding related to navigating or determining routes within an environment or system.
-
B.
navigationLandmarkFor
Indicates that one entity serves as a reference point or guide used to navigate to or within another entity.
-
C.
locationDirection
Indicates that one entity is located in a specific directional relation (such as north, south, above, or below) relative to another entity.
-
D.
mapLocation
Indicates a relationship where an entity is associated with a specific position or area on a map.
-
E.
notableCoPilotOnRoute
Indicates that an entity serves as a particularly recognized or distinguished co-pilot on a specific route.
- 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_69f76db20dac8190b1e8d0ca4dc1d59f |
completed | May 3, 2026, 3:45 p.m. |
| NER | Named-entity recognition | batch_69f77ffa6b68819090257fed3802c239 |
completed | May 3, 2026, 5:03 p.m. |
| PD | Predicate disambiguation | batch_69f7795978c481909e152cd1bd02dd07 |
completed | May 3, 2026, 4:35 p.m. |
| PDg | Predicate description generation | batch_69f77ff804f08190b431a31e6179ace4 |
completed | May 3, 2026, 5:03 p.m. |
Created at: May 3, 2026, 3:59 p.m.