Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Medical Domain

Benchmarks

Task NameDataset NameSOTA ResultTrend
Medical Knowledge Question AnsweringMedical Domain (MedQA, MMLU, MedMCQA) (test)
MedQA Score92.5
45
Spearman rank correlation with retrievalMedical Domain ROCO, MIMIC-CXR
Spearman Rho0.969
9
Entity RecognitionMedical Domain 8 held-out sets
EER5.84
8
SummarizationMedical domain (100 random samples)
Coherence3.3
4
Medical ReasoningMedical domain (val)
Correctness3.8
2
Showing 5 of 5 rows