Triple

T5135634
Position Surface form Disambiguated ID Type / Status
Subject Grendel's mother E115812 entity
Predicate attacks P34832 FINISHED
Object Heorot E493752 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: Heorot | Statement: [Grendel's mother, attacks, Heorot]
NED1 Entity disambiguation (via context triple) gpt-5-mini-2025-08-07
Target entity: Heorot
Context triple: [Grendel's mother, attacks, Heorot]
  • A. Heorot chosen
    Heorot is the great mead hall in the Old English epic Beowulf, famed as King Hrothgar’s royal hall and the site of Grendel’s attacks.
  • B. Halldon
    Halldon is a fictional continent on the planet Skaro in the Doctor Who universe, known as part of the Daleks’ homeworld geography.
  • C. Håkons Hall
    Håkons Hall is a large multi-purpose indoor arena in Lillehammer, Norway, best known for hosting ice hockey events during the 1994 Winter Olympics.
  • D. Throne Hall
    The Throne Hall is a grand ceremonial chamber within Gatchina Palace, historically used for imperial receptions and official court functions.
  • E. Loevestein Castle
    Loevestein Castle is a medieval Dutch fortress in the Netherlands, best known as a state prison and the site of Hugo Grotius’s famous escape in a book chest.
  • 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_69bd44459a988190a772a5c2ec6a1965 completed March 20, 2026, 12:57 p.m.
NER Named-entity recognition batch_69bd785069108190bf9cfdc7d962d43f completed March 20, 2026, 4:39 p.m.
NED1 Entity disambiguation (via context triple) batch_69befe51fe988190afcba16381b33043 completed March 21, 2026, 8:23 p.m.
Created at: March 20, 2026, 1:43 p.m.