Triple

T16097481
Position Surface form Disambiguated ID Type / Status
Subject Jeopardy! E390524 entity
Predicate creator P184 FINISHED
Object Merv Griffin E202328 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: Merv Griffin | Statement: [Jeopardy!, creator, Merv Griffin]
NED1 Entity disambiguation (via context triple) gpt-5-mini-2025-08-07
Target entity: Merv Griffin
Context triple: [Jeopardy!, creator, Merv Griffin]
  • A. Merv Griffin chosen
    Merv Griffin was an American television host, singer, and media mogul best known for creating the iconic game shows "Jeopardy!" and "Wheel of Fortune."
  • B. Mike Douglas
    Mike Douglas was an American singer and television talk show host best known for "The Mike Douglas Show," a popular daytime program that ran from the 1960s through the early 1980s.
  • C. Chuck Barris
    Chuck Barris was an American television producer and game show creator best known for developing and hosting "The Gong Show" and for his controversial, semi-fictional memoir "Confessions of a Dangerous Mind."
  • D. Don Martin
    Don Martin was a screenwriter known for his work on mid-20th-century American films, including the Western "Arrow in the Dust."
  • E. Don Martin
    Don Martin was an American cartoonist best known for his absurd, slapstick-heavy comics and uniquely stylized characters in Mad magazine.
  • 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_69d87f198bc48190a8b7e53ca15b7ead completed April 10, 2026, 4:39 a.m.
NER Named-entity recognition batch_69e1ff6551a48190afb7e0c61e22b541 completed April 17, 2026, 9:37 a.m.
NED1 Entity disambiguation (via context triple) batch_69fff79a96d08190af69cbb18037f66e completed May 10, 2026, 3:12 a.m.
Created at: April 10, 2026, 4:59 a.m.