johnsnowlabs/JSL-MedMNX-7B

Reported on 3 benchmarks across 3 tasks · 1 paper

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Methodology3 results

Few-Shot LearningonMedConceptsQA
Accuracy· 2024-05-12
25.627
best: 61.911 (gpt-4-0125-preview)
MedConceptsQA: Open Source Medical Concepts QA Benchmark arXiv:2405.07348
Meta-LearningonMedConceptsQA
Accuracy· 2024-05-12
25.627
best: 61.911 (gpt-4-0125-preview)
MedConceptsQA: Open Source Medical Concepts QA Benchmark arXiv:2405.07348
Zero-Shot LearningonMedConceptsQA
Accuracy
24.427
best: 52.489 (gpt-4-0125-preview)