Triple

T5075471
Position Surface form Disambiguated ID Type / Status
Subject Bhima E114383 entity
Predicate killed P4646 FINISHED
Object Kichaka E473858 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: Kichaka | Statement: [Bhima, killed, Kichaka]
NED1 Entity disambiguation (via context triple) gpt-5-mini-2025-08-07
Target entity: Kichaka
Context triple: [Bhima, killed, Kichaka]
  • A. Kichaka chosen
    Kichaka is a powerful and arrogant commander in the Mahabharata who serves as the main antagonist during the Pandavas’ incognito stay in King Virata’s kingdom.
  • B. Kokane
    Kokane is an American rapper and singer known for his distinctive G-funk style and frequent collaborations with West Coast hip hop artists.
  • C. Hacha-Kekan
    Hacha-Kekan is a traditional cultural festival of the Karbi people that showcases their indigenous rituals, music, dance, and communal celebrations.
  • D. Harku
    Harku is a small settlement in northern Estonia located within Harku Parish, near the capital city of Tallinn.
  • E. Kodo
    Kodo is the main lecture and assembly hall of the historic Yakushi-ji Buddhist temple in Nara, Japan.
  • 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_69bd443dbf908190a9401e9c2dc7bd7d completed March 20, 2026, 12:57 p.m.
NER Named-entity recognition batch_69bd74d2243481908c1ae62f7123c4e9 completed March 20, 2026, 4:24 p.m.
NED1 Entity disambiguation (via context triple) batch_69beb121fa388190ab3909d20f014b6f completed March 21, 2026, 2:54 p.m.
Created at: March 20, 2026, 1:39 p.m.