Observed arrival · 2026-11-02
BenchToward’s open clinical-AI scoreboards
An independent project publishes medical-AI evaluations on Korean clinical datasets, including scores, item-level predictions, and model versions.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- Clinical AI researchers and benchmark users
- Worth noticing
- Results include item-level predictions and exact model versions; results use CC BY 4.0, while each dataset keeps its own license.
Field notes
The scoreboards pair model results with reference baselines, including simple priors and named author scripts, making the comparisons easier to interpret from the page. The displayed tasks cover several data types, and the project reports 86 reuses; a separate benchmark builder is still marked as building, and the site labels itself Prototype (P0).
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
✦PrettyNotable craft visible
●ProPolished or operationally mature
◎NicheUnusually specific use
One card from the complete issue