AI2 Reasoning Challenge (ARC)
Question AnsweringReading ComprehensionEnglishBenchmark
AI2 Reasoning Challenge (ARC) is a question answering-focused benchmark dataset in English that provides 7,787 labeled examples distributed in JSON, CSV format.
📊 This dataset is used as an LLM benchmark. See model leaderboards →
About AI2 Reasoning Challenge (ARC)
Dataset contains 7,787 genuine grade-school level, multiple-choice science questions.
Details
- Task
- Question Answering, Reading Comprehension
- Language
- English
- Format
- JSON, CSV
- Rows / instances
- 7,787
- Creator
- Clark et al.
- Year
- 2018