Medicine
Real-POCQi
Real-POCQi: blinded specialty-matched physician pairwise evaluation of clinical AI systems (GPT-5.5, Gemini 3.1 Pro, Claude Opus 4.8, OpenEvidence) on 620 real point-of-care queries, rated on five dimensions.
3,100items
4subjects
CC-BY-4.0 (dataset); MIT (analysis code)license
medicinedomain
preferencedomain
textmodality
item-level responses released
Saturation status: No
Response matrix
Loading response matrix…
lowhighUnobserved
Scale: Grades: 0, 0.5, 1