🧪 Test?View on arXiv
MedJev: Evaluating Zero-Shot System-One Decision Making in Medicine
Not specified in the provided content
zero-shotclinical decision supportmachine learninglocal calibration
2610.36006
Builder Relevance
1h ago80%
Abstract
MedJev is an open System One model for clinical decision support that achieves competitive performance on medical benchmarks while running locally on a laptop.
Reality Card
Core Claim
MedJev achieves accuracies approaching GPT-4 performance on medical benchmarks while maintaining low computational costs and efficient local calibration.
Method / Result
MedJev achieves an AUROC of 0.926 for In Hospital Mortality predictions.
Limitations
Patient level predictions exhibit substantial miscalibration, particularly in In Hospital Mortality.
Paper to code
Verified implementation resources so builders can test the paper’s claims instead of stopping at the abstract.
No verified implementation link has been attached yet. AIBuzzHub will keep this panel separate from unverified search results.