Triple

T23014845
Position Surface form Disambiguated ID Type / Status
Subject Grunewald E573002 entity
Predicate hasLandmark P105 FINISHED
Object AVUS NE NERFINISHED

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: AVUS | Statement: [Grunewald, hasLandmark, AVUS]
NED1 Entity disambiguation (via context triple) gpt-5-mini-2025-08-07
Target entity: AVUS
Context triple: [Grunewald, hasLandmark, AVUS]
  • A. AVUS chosen
    AVUS is a historic German motor racing circuit and former public road in Berlin, renowned as one of the world’s earliest purpose-built race tracks.
  • B. AVUM
    AVUM is the liquid-propellant upper stage of the European Vega small-lift launch vehicle, responsible for precise orbital insertion and mission completion.
  • C. Uesen
    Uesen is a district or locality within the town of Achim in Lower Saxony, Germany.
  • D. AUT
    AUT is a leading Iranian engineering and technology university, widely recognized for its strong research output and rigorous academic programs.
  • E. AUT
    AUT is a major New Zealand university based in Auckland, known for its focus on applied research, innovation, and industry-aligned education.
  • F. None of above.
  • G. Unsure - the case is ambiguous/there is not enough information to decide.

Provenance (2 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_69e245b764cc8190a51be76f1d9611e1 completed April 17, 2026, 2:37 p.m.
NER Named-entity recognition batch_69f183e3c0e08190a7ac747b056ec3ca completed April 29, 2026, 4:06 a.m.
Created at: April 17, 2026, 3:51 p.m.