Dataset · mirobody
Medication reconciliation benchmark for LLMs
mirobody/MedTEMPO-Bench
A longitudinal benchmark for verifying medication regimen consistency during hospital admissions.
What's inside
Contains inpatient medical records in English and Chinese reflecting medication regimen changes (discontinued, substituted, or dosage-adjusted drugs) upon patient readmission. The agent acts as the attending physician.
What you can do with it
- Evaluating clinical reasoning in medical LLM agents
- Testing prescription and discharge reconciliation systems
- Tasks
- Question answering
- Languages
- English, Chinese
- Size
- 1K–10K rows
- Downloads this month
- 992
- Likes
- 18
How to load it
from datasets import load_dataset
ds = load_dataset("mirobody/MedTEMPO-Bench")