Triple
T22593444
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Mark Levin |
E565008
|
entity |
| Predicate | name |
P16
|
FINISHED |
| Object | Mark Levin |
—
|
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: Mark Levin | Statement: [Mark Levin, name, Mark Levin]
NED1
Entity disambiguation (via context triple)
gpt-5-mini-2025-08-07
Target entity: Mark Levin Context triple: [Mark Levin, name, Mark Levin]
-
A.
Mark Levin
chosen
Mark Levin is a conservative American lawyer, author, and talk radio and television host known for his outspoken commentary on U.S. politics and constitutional issues.
-
B.
Mark Levin
Mark Levin is an American screenwriter and producer known for co-writing family-oriented films and television projects, including the sci-fi adventure movie "The Adam Project."
-
C.
John Batchelor
John Batchelor is an Australian actor known for his work in film, television, and theatre.
-
D.
Larry Klayman
Larry Klayman is an American lawyer and conservative activist best known as the founder of the watchdog organization Judicial Watch.
-
E.
Michael Allen Jones
Michael Allen Jones is an American rapper better known by his stage name Mike Jones, who gained mainstream popularity in the mid-2000s with hits like "Still Tippin'" and "Back Then."
- F. None of above.
- G. Unsure - the case is ambiguous/there is not enough information to decide.
Provenance (2 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_69e245836014819091b91ed3074742a3 |
completed | April 17, 2026, 2:36 p.m. |
| NER | Named-entity recognition | batch_69f16163cb248190b377b110d80a6730 |
completed | April 29, 2026, 1:39 a.m. |
Created at: April 17, 2026, 2:49 p.m.