ctxdexSynthetic model data for learning

Benchmark family

AI2 Reasoning Challenge

A family groups the versions of one benchmark. It is not evaluated and has no applicable models of its own; each of its 3 definitions is resolved separately.

General reasoningKnowledgeScience reasoning

Definitions

Compare versions →

Newest first, following the recorded successor chain. Each definition is resolved against models separately.

  1. BenchmarkAI2 Reasoning Challenge (ARC)
    ActiveNewest in family
    Scale
    7,787 questions
    Primary metric
    Accuracy
    Applicable models
    30
  2. BenchmarkARC-Challenge
    Active
    Subset of AI2 Reasoning Challenge (ARC)
    Scale
    Not recorded
    Primary metric
    Accuracy
    Applicable models
    30
  3. BenchmarkARC-DA
    Active
    Derived from AI2 Reasoning Challenge (ARC)
    Scale
    Not recorded
    Primary metric
    Exact Match
    Applicable models
    30

Family methodology

No family-level methodology recorded.

Across definitions

Tasks
Grade-school science multiple-choice Q&A
Sources
None recorded at family level
Last reviewed
2026-09-22

Applicability does not mean evaluated; benchmark result/model score data is outside this UI contract.