RaindropA practical guide to choosing decision modelsGitHub ↗

Relations · question 63

Back to Relations · All capabilities

Question ID: bbh/264dee8d07c30728dbfc
Source: BIG-Bench Hard selected tasks
Input format: native choices

Explore this question’s altered versions

Question

The following paragraphs each describe a set of seven objects arranged in a fixed order. The statements are logically consistent within each paragraph. In a golf tournament, there were seven golfers: Joe, Dan, Ada, Amy, Rob, Mya, and Mel. Ada finished below Amy. Joe finished below Dan. Dan finished below Ada. Mel finished third-to-last. Amy finished third. Rob finished below Mya.

Answer choices

  • A: Joe finished last
  • B: Dan finished last
  • C: Ada finished last
  • D: Amy finished last
  • E: Rob finished last
  • F: Mya finished last
  • G: Mel finished last

Reference answer: Joe finished last (A)

Exact evaluated input as JSON
{
  "state": "",
  "question": "The following paragraphs each describe a set of seven objects arranged in a fixed order. The statements are logically consistent within each paragraph. In a golf tournament, there were seven golfers: Joe, Dan, Ada, Amy, Rob, Mya, and Mel. Ada finished below Amy. Joe finished below Dan. Dan finished below Ada. Mel finished third-to-last. Amy finished third. Rob finished below Mya.",
  "choices": {
    "A": "Joe finished last",
    "B": "Dan finished last",
    "C": "Ada finished last",
    "D": "Amy finished last",
    "E": "Rob finished last",
    "F": "Mya finished last",
    "G": "Mel finished last"
  }
}

Recorded model answers

Model Answer Outcome
Astra Joe finished last Correct
Opus 5.5 Joe finished last Correct
GLiDE v1.2 Joe finished last Correct
Decider v1.1 Joe finished last Correct
Jev Joe finished last Correct
Microsoft Decision-1 Joe finished last Correct
Clef Ada finished last Incorrect
Luna Decisions Joe finished last Correct
Clef Flash Joe finished last Correct
Strands Decider Mel finished last Incorrect
GLiNER2.5 Decide Ada finished last Incorrect
Laya Dan finished last Incorrect
CalDec v1 Mel finished last Incorrect

Source and sharing

Attribution: Mirac Suzgun and BIG-Bench Hard authors; BIG-bench contributors.
Terms: MIT AND 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": "revision-15-abilities/logical_deduction_seven_objects.json",
    "record": "64",
    "kind": "source_file_record"
  },
  "source_revision": "9ee07bd481feebf959a6b59d61ea57bdcf30964d",
  "input_hash": "a71c954372d0624ef081e69016770ec4d7b0a31e3d693af997055d88b260da1e"
}

Type to search questions, variants and documentation.