What it does
On its jev-by-example cases two small models diverged on 6 of 28 application decisions, including one where a double would retry a write that may already have succeeded.
Maker-reported (not independently measured by JevMade): 6 of 28 application decisions diverged between two small models (jev-by-example suite) · reported suites: jevbench-hard, kev-decision-v2-300, jev-by-example
- Primitives
- choice, score, noul
- Platform
- JavaScript (Node)
- Added
- Project created
Source checked 2026-09-24 — opened the GitHub repository directly.