FirsthandHealth
medRxiv PreprintsInternational17 November 2026

Evaluation of System One Models on Biomedical Texts

This is an official announcement record

Firsthand records what medRxiv Preprints announced and links to the original. The wording below is theirs, not ours.

Objective To compare the performance and cost of System One models (SOMs) with general-purpose large language models (LLMs) on biomedical tasks and evaluate confidence-based routing. Materials and Methods We evaluated three SOMs and three LLMs across seven datasets and eight endpoints covering question answering, error detection, structured extraction, and clinical-note assessment. Outcomes included task-specific performance, recorded inference cost and latency, and confidence calibration. Calibration data determined routing schemes and confidence thresholds to evaluate test predictions, with
Source: medRxiv Preprints
Read the official announcement

Opens www.medrxiv.org

More from medRxiv Preprints

This content is for informational purposes only and is not medical advice. It is not intended to diagnose, treat, cure, or prevent any disease. Consult a healthcare professional before starting any supplement, treatment, or program: especially if you are pregnant, nursing, taking medication, or managing a health condition.