Triple

T11235335
Position Surface form Disambiguated ID Type / Status
Subject Bitter Moon E265927 entity
Predicate screenwriter P2831 FINISHED
Object Gérard Brach E311880 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: Gérard Brach | Statement: [Bitter Moon, screenwriter, Gérard Brach]
NED1 Entity disambiguation (via context triple) gpt-5-mini-2025-08-07
Target entity: Gérard Brach
Context triple: [Bitter Moon, screenwriter, Gérard Brach]
  • A. Gérard Brach chosen
    Gérard Brach was a French screenwriter and frequent Roman Polanski collaborator known for his work on numerous acclaimed European films.
  • B. Alain Glavieux
    Alain Glavieux was a French engineer and information theorist best known as a co-inventor of turbo codes, a breakthrough in error-correcting coding that revolutionized digital communications.
  • C. Michel Andrault
    Michel Andrault was a prominent French architect known for his influential large-scale housing and urban development projects in the late 20th century.
  • D. Pierre Manent
    Pierre Manent is a French political philosopher known for his work on liberalism, democracy, and the intellectual history of modern Europe.
  • E. René Le Hénaff
    René Le Hénaff was a French film editor and director known for his work on numerous classic French films in the early to mid-20th century.
  • 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_69d6aac656d48190b275efaa7d6074ee completed April 8, 2026, 7:21 p.m.
NER Named-entity recognition batch_69d7e904cf888190826fc964f76b5cb2 completed April 9, 2026, 5:59 p.m.
NED1 Entity disambiguation (via context triple) batch_69e6e6e19c9c81909d114cf9bd0e2f84 completed April 21, 2026, 2:54 a.m.
Created at: April 8, 2026, 9:30 p.m.