Triple

T15700470
Position Surface form Disambiguated ID Type / Status
Subject Joe Christmas E380578 entity
Predicate isKilledBy P4646 FINISHED
Object Percy Grimm E1171970 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: Percy Grimm | Statement: [Joe Christmas, isKilledBy, Percy Grimm]
NED1 Entity disambiguation (via context triple) gpt-5-mini-2025-08-07
Target entity: Percy Grimm
Context triple: [Joe Christmas, isKilledBy, Percy Grimm]
  • A. Percy Grimm chosen
    Percy Grimm is a fanatically nationalistic World War I veteran and self-appointed vigilante in William Faulkner’s novel "Light in August," known for his obsessive pursuit and brutal killing of Joe Christmas.
  • B. John "Reaper" Grimm
    John "Reaper" Grimm is the tough, battle-hardened space marine protagonist of the 2005 sci-fi horror film "Doom," known for leading a squad against mutated creatures on Mars.
  • C. Ian Grimm
    Ian Grimm is the egotistical yet charismatic creative director at the center of the video game studio comedy series "Mythic Quest."
  • D. Tuffnut Thorston
    Tuffnut Thorston is a boisterous, reckless Viking twin and dragon rider known for his comic relief role in the How to Train Your Dragon franchise.
  • E. Percy LeSueur
    Percy LeSueur was a prominent early 20th-century Canadian ice hockey goaltender and later coach and manager, recognized as a Hall of Famer for his contributions to the sport.
  • 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_69d86d99e860819094b6957cde470f2c completed April 10, 2026, 3:25 a.m.
NER Named-entity recognition batch_69e04f6d71308190971c10c599da9645 completed April 16, 2026, 2:54 a.m.
NED1 Entity disambiguation (via context triple) batch_69ff82f05d648190a0c73b60dc027287 completed May 9, 2026, 6:54 p.m.
Created at: April 10, 2026, 4:45 a.m.