Benchmark RadarResearchQA
ResearchQA is a benchmark of 6,211 single-paper question-answer pairs from 494 open-access papers across eight domains, designed for citation-grounded evaluation with multiple valid supporting passages and grounded refusal.
- Released
- 2026-07-13
- Readiness
- Paper only
- Primary field
- General AI
Why it matters
Existing evaluation methods often fail to detect whether answers are supported by verifiable citations. ResearchQA provides a standardized way to assess citation accuracy and groundedness, separating systems more clearly than LLM-evaluator scores.
Motivation
Large language models are increasingly used to assist scientific reading, but existing evaluation methods often fail to detect whether answers are supported by verifiable citations.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.