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: Vubr has a direct effect on wibl and uvzi. Vubr is unobserved. For those who are not zuph, the probability of uvzi is 26%. Zuph has a direct effect on wibl and uvzi. Wibl has a direct effect on uvzi. The overall probability of zuph is 56%. For those who are zuph, the probability of uvzi is 46%.

Question

Is uvzi less likely than not uvzi overall?

Answer choices

  • yes. Yes
  • no. No

Reference answer: yes.

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: Vubr has a direct effect on wibl and uvzi. Vubr is unobserved. For those who are not zuph, the probability of uvzi is 26%. Zuph has a direct effect on wibl and uvzi.\nWibl has a direct effect on uvzi. The overall probability of zuph is 56%. For those who are zuph, the probability of uvzi is 46%.",
  "question": "Is uvzi less likely than not uvzi overall?",
  "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 yes — Yes yes — Yes Same answer
Decider v1.1 yes — Yes yes — Yes Same answer
Jev yes — Yes yes — Yes Same answer
Microsoft Decision-1 no — No no — No Same answer
Clef yes — Yes yes — Yes Same answer
Luna Decisions no — No yes — Yes Different answer
Clef Flash yes — Yes yes — Yes 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 requested source unit permutation. Only the exact matched comparison sets contribute to the reported ordering contrasts.

Identifiers and recorded conditions

Variant ID: cladder/665da985e08fe737ecdd/presentation_order/shuffled_presentation/clean-bare_causal_premise_order_full_0034

Input fingerprint: 2c679c9d5898e579362fe9985480d3959a6eee6075b55844ae3b425ef668c209

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.