Triple

T5293912
Position Surface form Disambiguated ID Type / Status
Subject Blade Runner 2049 E119808 entity
Predicate producer P490 FINISHED
Object Andrew A. Kosove E395694 NE 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: Andrew A. Kosove | Statement: [Blade Runner 2049, producer, Andrew A. Kosove]
NED1 Entity disambiguation (via context triple) gpt-5-mini-2025-08-07
Target entity: Andrew A. Kosove
Context triple: [Blade Runner 2049, producer, Andrew A. Kosove]
  • A. Andrew A. Kosove chosen
    Andrew A. Kosove is an American film producer and co-founder of Alcon Entertainment, known for backing major Hollywood features across genres.
  • B. Ronald M. Cohen
    Ronald M. Cohen was an American screenwriter known for his work on politically charged and socially conscious films and television projects.
  • C. Russell Gewirtz
    Russell Gewirtz is an American screenwriter best known for writing the crime thriller film "Inside Man."
  • D. Michael S. Chernuchin
    Michael S. Chernuchin is an American television writer and producer best known for his extensive work on the "Law & Order" franchise.
  • E. Andrew Weisblum
    Andrew Weisblum is an American film editor known for his work on major feature films, including collaborations with directors like Darren Aronofsky and Wes Anderson.
  • F. None of above.
  • G. Unsure - the case is ambiguous/there is not enough information to decide.

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_69bd446f22b88190b6a47fb91c68a3e7 completed March 20, 2026, 12:58 p.m.
NER Named-entity recognition batch_69bd84f034f081908027a43120b6e122 completed March 20, 2026, 5:33 p.m.
NED1 Entity disambiguation (via context triple) batch_69bfc60964c08190bcb128946e121bc9 completed March 22, 2026, 10:35 a.m.
Created at: March 20, 2026, 1:52 p.m.