RaindropA practical guide to choosing decision modelsGitHub ↗

Learning and induction · question 11

Back to Learning and induction · All capabilities

Question ID: list_functions/155ff0b3e37b0628b8bc
Source: List Functions
Input format: adapted choices

Explore this question’s altered versions

Context

[{"input": "[36, 1, 44, 15, 20, 4, 62, 82, 21]", "output": "[82]"}, {"input": "[56, 30, 9, 16, 15, 20, 8, 98, 5]", "output": "[98, 98, 98, 98, 98]"}, {"input": "[8, 3, 84, 6]", "output": "[84, 84, 84]"}, {"input": "[10, 7, 80, 8]", "output": "[80, 80, 80, 80, 80, 80, 80]"}, {"input": "[3, 2, 77]", "output": "[77, 77]"}, {"input": "[89, 4]", "output": "[89, 89, 89, 89]"}, {"input": "[8]", "output": "[8, 8, 8, 8, 8, 8, 8, 8]"}, {"input": "[71, 85, 15, 7, 3]", "output": "[85, 85, 85]"}]

Question

Apply the same inferred rule to this new input: [18, 35, 5, 8, 57, 3, 70, 22, 11, 6]

Answer choices

  • A: [24, 46, 27]
  • B: [18, 5, 57]
  • C: [18, 8, 70]
  • D: [70, 70, 70]

Reference answer: [70, 70, 70] (D)

Exact evaluated input as JSON
{
  "state": "[{\"input\": \"[36, 1, 44, 15, 20, 4, 62, 82, 21]\", \"output\": \"[82]\"}, {\"input\": \"[56, 30, 9, 16, 15, 20, 8, 98, 5]\", \"output\": \"[98, 98, 98, 98, 98]\"}, {\"input\": \"[8, 3, 84, 6]\", \"output\": \"[84, 84, 84]\"}, {\"input\": \"[10, 7, 80, 8]\", \"output\": \"[80, 80, 80, 80, 80, 80, 80]\"}, {\"input\": \"[3, 2, 77]\", \"output\": \"[77, 77]\"}, {\"input\": \"[89, 4]\", \"output\": \"[89, 89, 89, 89]\"}, {\"input\": \"[8]\", \"output\": \"[8, 8, 8, 8, 8, 8, 8, 8]\"}, {\"input\": \"[71, 85, 15, 7, 3]\", \"output\": \"[85, 85, 85]\"}]",
  "question": "Apply the same inferred rule to this new input: [18, 35, 5, 8, 57, 3, 70, 22, 11, 6]",
  "choices": {
    "A": "[24, 46, 27]",
    "B": "[18, 5, 57]",
    "C": "[18, 8, 70]",
    "D": "[70, 70, 70]"
  }
}

Recorded model answers

Model Answer Outcome
Astra [70, 70, 70] Correct
Opus 5.5 [70, 70, 70] Correct
GLiDE v1.2 [70, 70, 70] Correct
Decider v1.1 [70, 70, 70] Correct
Jev [70, 70, 70] Correct
Microsoft Decision-1 [70, 70, 70] Correct
Clef [70, 70, 70] Correct
Luna Decisions [70, 70, 70] Correct
Clef Flash [70, 70, 70] Correct
Strands Decider [70, 70, 70] Correct
GLiNER2.5 Decide [18, 5, 57] Incorrect
Laya [18, 8, 70] Incorrect
CalDec v1 [18, 5, 57] Incorrect

Source and sharing

Licensed upstream task: c111. All eight supplied examples and the evaluated test input/output match the original BIG-bench data.

Attribution: Joshua S. Rule, Steven T. Piantadosi, Joshua B. Tenenbaum; BIG-bench contributors; Linlu Qiu et al. (selection).
Terms: Apache-2.0. Source sharing decision.
The capability assignment and choice-format normalization are part of this experiment. Source data retain their own terms.

Exact source locator and input fingerprint
{
  "source_locator": {
    "path": "data/list_function.jsonl",
    "record": "110/5",
    "kind": "source_file_record"
  },
  "source_revision": "0eae21f3d3014b92be34c80b23c9120a7b15a51f",
  "input_hash": "dad396ce54f055fe90be6d8c4e83cd36f0a3510d7b0a7224454e6e04f86f6b67"
}

Type to search questions, variants and documentation.