Benchmark RadarAI BENCHMARK PROFILE
APS-Bench
APS-Bench is a 50-question QA dataset with auditable gold answers for evaluating retrieval-augmented generation over scientific facility operations knowledge.
- Released
- 2026-07-27
- Readiness
- Paper only
- Primary field
- Science & Research
Why it matters
It addresses the need for evaluating RAG systems on niche institutional knowledge, but its small scale and facility-specific scope limit broader utility.
Motivation
Scientific user facilities accumulate decades of operational knowledge that no single search index covers: electronic logbooks, technical documents, internal wikis, operations chat messages, maintenance records, and live control-system data.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.