Benchmarks
How each medbench.kr benchmark is made: the published evaluation it adopts, the open SNUH data it runs on, and where it stands. Built by the medbench.kr benchmark builder; watch builds on Live.
Code on GitHub: medbench-kr/benchmarks (Apache-2.0, problem reports as issues). Data on KHDP, the data platform of Seoul National University Hospital: each benchmark links its dataset there, and datasets medbench.kr builds are uploaded to KHDP with a DOI.
All 9 Building 4 In review 0 Published 5
| Benchmark | Data | Adopted from | Status | |
|---|---|---|---|---|
| Intraoperative hypotension, the authors' unbiased dataset vitaldb-ioh@1 |
VitalDB | hyunlimy/intraop_hypotension (MIT) | building | Build 50.5% |
| In-ICU cardiac arrest from ECG and PPG scope-arrest@0 |
SCOPE (SNUH ICU) | medbench.kr definition | building | Protocol |
| Text-to-SQL over OMOP CDM synthetic-snuh-text2sql@0 |
Synthetic-SNUH v4 | EHRSQL question taxonomy and execution-accuracy scorer | building | Build |
| Korean surgical robot voice commands kor-surgical-robot-voice-command@0 |
Korean Surgical Robot Voice-Command Benchmark (KHDP) | the authors' benchmark on KHDP (doi:10.23421/2026.03507) | building | Baselines |