RaindropA practical guide to choosing decision modelsGitHub ↗

Irrelevant details · Add a larger amount of irrelevant detail variant

Original question · Other versions of this question · All variants

This version keeps the original reference answer. Its context, question instructions or answer choices were changed as described below.

Changed fields: context.

How the edits were constructed

Context

[Gloria] went to the store with her sister [Stephanie] [Jason] took his daughter [Stephanie] to the park that afternoon to play. [Gloria] went to her brother [Jeff]'s birthday party [Jeremy] and his sister, [Gloria] love it when their grandmother visits because she spoils them. She is coming this week to watch them while their parents are out of town. [Gloria] and [Jeff] are siblings who are enjoying lunch together. The lunch plates were white. The cups were blue. The serving tray was oval. The napkins were paper. The table top was wooden. The floor was stone.

Question

What is Jeremy’s relationship to Jason?

Answer choices

  • aunt. aunt
  • brother. brother
  • daughter. daughter
  • daughter-in-law. daughter-in-law
  • father. father
  • father-in-law. father-in-law
  • granddaughter. granddaughter
  • grandfather. grandfather
  • grandmother. grandmother
  • grandson. grandson
  • mother. mother
  • mother-in-law. mother-in-law
  • nephew. nephew
  • niece. niece
  • sister. sister
  • son. son
  • son-in-law. son-in-law
  • uncle. uncle

Reference answer: son.

Exact evaluated input as JSON
{
  "state": "[Gloria] went to the store with her sister [Stephanie] [Jason] took his daughter [Stephanie] to the park that afternoon to play. [Gloria] went to her brother [Jeff]'s birthday party [Jeremy] and his sister, [Gloria] love it when their grandmother visits because she spoils them. She is coming this week to watch them while their parents are out of town. [Gloria] and [Jeff] are siblings who are enjoying lunch together.\nThe lunch plates were white.\nThe cups were blue.\nThe serving tray was oval.\nThe napkins were paper.\nThe table top was wooden.\nThe floor was stone.",
  "question": "What is Jeremy’s relationship to Jason?",
  "choices": {
    "aunt": "aunt",
    "brother": "brother",
    "daughter": "daughter",
    "daughter-in-law": "daughter-in-law",
    "father": "father",
    "father-in-law": "father-in-law",
    "granddaughter": "granddaughter",
    "grandfather": "grandfather",
    "grandmother": "grandmother",
    "grandson": "grandson",
    "mother": "mother",
    "mother-in-law": "mother-in-law",
    "nephew": "nephew",
    "niece": "niece",
    "sister": "sister",
    "son": "son",
    "son-in-law": "son-in-law",
    "uncle": "uncle"
  }
}

Recorded model answers

The original column uses the control response paired with this variant. Missing responses and inputs outside a model’s limit remain visible.

Model Original answer Variant answer Change
GLiDE v1.2 son — son son — son Same answer
Decider v1.1 son — son son — son Same answer
Jev son — son son — son Same answer
Microsoft Decision-1 nephew — nephew nephew — nephew Same answer
Clef uncle — uncle uncle — uncle Same answer
Luna Decisions brother — brother Refused —
Clef Flash uncle — uncle uncle — uncle Same answer
Strands Decider sister — sister sister — sister Same answer
GLiNER2.5 Decide grandmother — grandmother grandmother — grandmother Same answer
Laya grandmother — grandmother grandmother — grandmother Same answer
CalDec v1 grandmother — grandmother grandmother — grandmother Same answer
Identifiers and recorded conditions

Variant ID: clutrr/28c1072a922ec496c68e/irrelevant_details/high/clean-remaining_kinship_details_full_0021

Input fingerprint: 670d8e0052148139bd86052a7e03920311de18fda1254e3880c09d61a5ebd118

Change Recorded version
Irrelevant details Add a larger amount of irrelevant detail

Source attribution and dataset terms are linked from the original question.

Type to search questions, variants and documentation.