Triple

T20033333
Position Surface form Disambiguated ID Type / Status
Subject Joshua Lawrence Chamberlain’s bayonet charge at Little Round Top E497193 entity
Predicate directionOfAttack P70449 FINISHED
Object downhill from Little Round Top LITERAL 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: downhill from Little Round Top | Statement: [Joshua Lawrence Chamberlain’s bayonet charge at Little Round Top, directionOfAttack, downhill from Little Round Top]
PD Predicate disambiguation gpt-5-mini-2025-08-07
Target predicate: directionOfAttack
Context triple: [Joshua Lawrence Chamberlain’s bayonet charge at Little Round Top, directionOfAttack, downhill from Little Round Top]
  • A. primaryAdversaryDirection
    Indicates the direction from an entity toward its main or most significant adversary.
  • B. attackedFrom chosen
    Indicates that one entity initiated an attack against another entity originating from a specific source location or position.
  • C. attackType
    Indicates the specific method, style, or category of attack used in an aggressive or hostile action between entities.
  • D. objectiveOfAttackers
    Indicates the target or goal that attackers aim to reach or affect through their attack.
  • E. directionOfEncirclement
    Indicates the directional orientation (e.g., clockwise or counterclockwise) in which one entity surrounds or encircles another.
  • F. None of above.

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_69da627278c88190babe4297a9df1236 completed April 11, 2026, 3:02 p.m.
NER Named-entity recognition batch_69e66293e3908190ade07ba9e09da803 completed April 20, 2026, 5:29 p.m.
PD Predicate disambiguation batch_69e54ce752748190a0a1ffddd0372271 completed April 19, 2026, 9:45 p.m.
Created at: April 11, 2026, 3:36 p.m.