Triple

T1940005
Position Surface form Disambiguated ID Type / Status
Subject Khoms District E41530 entity
Predicate capital P234 FINISHED
Object Khoms E167177 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: Khoms | Statement: [Khoms District, capital, Khoms]
NED1 Entity disambiguation (via context triple) gpt-5-mini-2025-08-07
Target entity: Khoms
Context triple: [Khoms District, capital, Khoms]
  • A. Khoms chosen
    Khoms is a coastal city in northwestern Libya known for its proximity to the ancient Roman archaeological site of Leptis Magna.
  • B. Sadras
    Sadras is a historic coastal town in Tamil Nadu, India, known for its Dutch-era fort and role as a former trading port on the Coromandel Coast.
  • C. Gorgan
    Gorgan is a city in northern Iran, serving as the capital of Golestan Province near the southeastern coast of the Caspian Sea.
  • D. Lalish
    Lalish is a sacred mountain valley in northern Iraq that serves as the principal pilgrimage site and spiritual center of the Yazidi faith.
  • E. Arak
    Arak is a major industrial city in central Iran known for its manufacturing sector and strategic economic importance.
  • 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_69a88649b24c819080047f26b6db2ded completed March 4, 2026, 7:21 p.m.
NER Named-entity recognition batch_69abb2ca385c8190ad0e4bafbb9b5e5c completed March 7, 2026, 5:08 a.m.
NED1 Entity disambiguation (via context triple) batch_69adfbb7616881908bfc82997538ca08 completed March 8, 2026, 10:44 p.m.
Created at: March 4, 2026, 7:36 p.m.