RaindropA practical guide to choosing decision modelsGitHub ↗

Learning and induction · question 61

Back to Learning and induction · All capabilities

Question ID: list_functions/26e6b047752745c67869
Source: List Functions
Input format: adapted choices

Explore this question’s altered versions

Context

[{"input": "[]", "output": "[]"}, {"input": "[2, 3, 9, 22, 0]", "output": "[2, 3, 9, 22, 22]"}, {"input": "[91, 2]", "output": "[89, 89]"}, {"input": "[54, 38]", "output": "[16, 16]"}, {"input": "[2, 95, 1, 93, 17, 4, 0, 68]", "output": "[2, 95, 1, 93, 17, 4, 95, 68]"}, {"input": "[9, 2, 80, 18, 47]", "output": "[9, 78, 78, 18, 47]"}, {"input": "[89, 4, 35, 76]", "output": "[85, 85, 35, 76]"}, {"input": "[88, 86, 66, 90, 5, 0, 51, 45]", "output": "[88, 86, 66, 90, 5, 90, 51, 45]"}]

Question

Apply the same inferred rule to this new input: [1, 0, 55]

Answer choices

  • A: [0, -1, 54]
  • B: [1, 55, 55]
  • C: [2, 0, 110]
  • D: [1, 55, 1]

Reference answer: [1, 55, 55] (B)

Exact evaluated input as JSON
{
  "state": "[{\"input\": \"[]\", \"output\": \"[]\"}, {\"input\": \"[2, 3, 9, 22, 0]\", \"output\": \"[2, 3, 9, 22, 22]\"}, {\"input\": \"[91, 2]\", \"output\": \"[89, 89]\"}, {\"input\": \"[54, 38]\", \"output\": \"[16, 16]\"}, {\"input\": \"[2, 95, 1, 93, 17, 4, 0, 68]\", \"output\": \"[2, 95, 1, 93, 17, 4, 95, 68]\"}, {\"input\": \"[9, 2, 80, 18, 47]\", \"output\": \"[9, 78, 78, 18, 47]\"}, {\"input\": \"[89, 4, 35, 76]\", \"output\": \"[85, 85, 35, 76]\"}, {\"input\": \"[88, 86, 66, 90, 5, 0, 51, 45]\", \"output\": \"[88, 86, 66, 90, 5, 90, 51, 45]\"}]",
  "question": "Apply the same inferred rule to this new input: [1, 0, 55]",
  "choices": {
    "A": "[0, -1, 54]",
    "B": "[1, 55, 55]",
    "C": "[2, 0, 110]",
    "D": "[1, 55, 1]"
  }
}

Recorded model answers

Model Answer Outcome
Astra [1, 55, 55] Correct
Opus 5.5 [1, 55, 55] Correct
GLiDE v1.2 [1, 55, 55] Correct
Decider v1.1 [1, 55, 55] Correct
Jev [1, 55, 55] Correct
Microsoft Decision-1 [1, 55, 55] Correct
Clef [1, 55, 55] Correct
Luna Decisions [1, 55, 55] Correct
Clef Flash [1, 55, 55] Correct
Strands Decider [1, 55, 55] Correct
GLiNER2.5 Decide [0, -1, 54] Incorrect
Laya [2, 0, 110] Incorrect
CalDec v1 [2, 0, 110] Incorrect

Source and sharing

Licensed upstream task: c144. 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": "143/0",
    "kind": "source_file_record"
  },
  "source_revision": "0eae21f3d3014b92be34c80b23c9120a7b15a51f",
  "input_hash": "44bae9691bb9cb22139aca898425c4ccdd9072e29285ecbc578c2828501f32ae"
}

Type to search questions, variants and documentation.