Stress test of TypeSafe AI's jev-1.13 System One model: limits, vagueness, calibration, adversarial, new patterns, LLM bake-off
Stars measure the whole repository, including work unrelated to Jev.
Documented records the linked documentation or source. JevHunt has not independently run or benchmarked this project.
- Documented limits are exact: 255 Choice options, 10 Score levels, ~32k state tokens, ~64k total. A 1-level Score and an empty state are accepted. Versioned model ids (`jev-1.13.0`, `jev-1.12`) are rejected; only `jev-latest` / `jev-preview` work.
Evidence commit: ed2adb7740d7f42b47d4390d4816b4dfdc7b8808
Discovered through: github-search.
How we review →