RaindropA practical guide to choosing decision modelsGitHub ↗

Learning and induction · question 55

Back to Learning and induction · All capabilities

Question ID: list_functions/02a749772367743d5cc2
Source: List Functions
Input format: adapted choices

Explore this question’s altered versions

Context

[{"input": "[0, 22, 6, 4, 9]", "output": "[9, 26, 12, 26, 9]"}, {"input": "[7, 37, 64]", "output": "[71, 74, 71]"}, {"input": "[3, 1, 8, 62, 66, 9]", "output": "[12, 67, 70, 70, 67, 12]"}, {"input": "[65, 84, 9, 3]", "output": "[68, 93, 93, 68]"}, {"input": "[0, 42, 8, 27, 86]", "output": "[86, 69, 16, 69, 86]"}, {"input": "[5, 9, 3, 0, 39, 58]", "output": "[63, 48, 3, 3, 48, 63]"}, {"input": "[76, 22, 25, 2, 1]", "output": "[77, 24, 50, 24, 77]"}, {"input": "[1, 18, 83]", "output": "[84, 36, 84]"}]

Question

Apply the same inferred rule to this new input: [8, 4, 45]

Answer choices

  • A: [4, 45, 8]
  • B: [45, 45, 45]
  • C: [45, 8, 4]
  • D: [53, 8, 53]

Reference answer: [53, 8, 53] (D)

Exact evaluated input as JSON
{
  "state": "[{\"input\": \"[0, 22, 6, 4, 9]\", \"output\": \"[9, 26, 12, 26, 9]\"}, {\"input\": \"[7, 37, 64]\", \"output\": \"[71, 74, 71]\"}, {\"input\": \"[3, 1, 8, 62, 66, 9]\", \"output\": \"[12, 67, 70, 70, 67, 12]\"}, {\"input\": \"[65, 84, 9, 3]\", \"output\": \"[68, 93, 93, 68]\"}, {\"input\": \"[0, 42, 8, 27, 86]\", \"output\": \"[86, 69, 16, 69, 86]\"}, {\"input\": \"[5, 9, 3, 0, 39, 58]\", \"output\": \"[63, 48, 3, 3, 48, 63]\"}, {\"input\": \"[76, 22, 25, 2, 1]\", \"output\": \"[77, 24, 50, 24, 77]\"}, {\"input\": \"[1, 18, 83]\", \"output\": \"[84, 36, 84]\"}]",
  "question": "Apply the same inferred rule to this new input: [8, 4, 45]",
  "choices": {
    "A": "[4, 45, 8]",
    "B": "[45, 45, 45]",
    "C": "[45, 8, 4]",
    "D": "[53, 8, 53]"
  }
}

Recorded model answers

Model Answer Outcome
Astra [53, 8, 53] Correct
Opus 5.5 [53, 8, 53] Correct
GLiDE v1.2 [53, 8, 53] Correct
Decider v1.1 [53, 8, 53] Correct
Jev [53, 8, 53] Correct
Microsoft Decision-1 [53, 8, 53] Correct
Clef [53, 8, 53] Correct
Luna Decisions [53, 8, 53] Correct
Clef Flash [45, 45, 45] Incorrect
Strands Decider [45, 45, 45] Incorrect
GLiNER2.5 Decide [4, 45, 8] Incorrect
Laya [4, 45, 8] Incorrect
CalDec v1 [4, 45, 8] Incorrect

Source and sharing

Licensed upstream task: c207. 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": "206/6",
    "kind": "source_file_record"
  },
  "source_revision": "0eae21f3d3014b92be34c80b23c9120a7b15a51f",
  "input_hash": "f9595cdb6f0b3a3eee94e7c95b431b676fc00d76c0a197204bad9c3d8bd4742f"
}

Type to search questions, variants and documentation.