RaindropA practical guide to choosing decision modelsGitHub ↗

Evidence order · Present the evidence in a shuffled order 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

Imagine a self-contained, hypothetical world with only the following conditions, and without any unmentioned factors or causal relationships: Jyka has a direct effect on lirg. hwax and jyka causes lirg. Hwax has a direct effect on jyka and lirg. We know that hwax causes jyka. We observed an individual is hwax.

Question

Would an individual is not lirg if jyka instead of not jyka?

Answer choices

  • yes. Yes
  • no. No

Reference answer: no.

Exact evaluated input as JSON
{
  "state": "Imagine a self-contained, hypothetical world with only the following conditions, and without any unmentioned factors or causal relationships: Jyka has a direct effect on lirg. hwax and jyka causes lirg. Hwax has a direct effect on jyka and lirg. We know that hwax causes jyka. We observed an individual is hwax.",
  "question": "Would an individual is not lirg if jyka instead of not jyka?",
  "choices": {
    "yes": "Yes",
    "no": "No"
  }
}

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 no — No no — No Same answer
Decider v1.1 no — No no — No Same answer
Jev no — No no — No Same answer
Microsoft Decision-1 no — No no — No Same answer
Clef no — No no — No Same answer
Luna Decisions no — No no — No Same answer
Clef Flash no — No no — No Same answer
Strands Decider no — No no — No Same answer
GLiNER2.5 Decide no — No no — No Same answer
Laya no — No no — No Same answer
CalDec v1 no — No no — No Same answer

Ordering comparison: literal units with boundary layout change. Only the exact matched comparison sets contribute to the reported ordering contrasts.

Identifiers and recorded conditions

Variant ID: cladder/634311372ed0498c91a9/presentation_order/shuffled_presentation/clean-mixed_causal_arithmetic_full_0002

Input fingerprint: b208f7fdf5a643e0754e0022ec572a18768f7ef62a1bace7ff9192d6c4f5da75

Change Recorded version
Evidence order Present the evidence in a shuffled order

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

Type to search questions, variants and documentation.