index / asp616848/better-jev-bench
better-jev-bench
A wide, multi-domain, multi-modal decision/classification benchmark corpus, built in the Jev typed-decision style — for training and evaluating System One models
Research and EvalsUnclearReview
Repository
Stars0
Language—
LicenseApache-2.0
Created2026-09-23
Last pushed2026-09-23
Officialno
Judgment
Genuine Jev project0.73
gate 0.5
Uses Jev at runtime0.15
Is a meta list0.22
Is a reimplementation0.25
Quality scales
Substance0.13 / 3
gate 0.5
Docs0.01 / 3
Novelty1.54 / 3
Composite score0.17
Category distribution
Research and Evals1.00
Integrations0.00
Applications0.00
Other0.00
Learning0.00
SDKs and Clients0.00
Agent and Dev Tooling0.00
Games and Simulation0.00
Assigned categoryResearch and Evals
Category probability1.00
Category confidence1.00
PatternUnclear
Pattern confidence0.79
Curation
StatusReview
Reasonsubstance 0.13 below 0.5
README pickno
Overridenone
Sources
- search:"system one" jev
- search:jev in:name,description,readme created:2026-09-23
Provenance
Modeljev-1.13.0
Question setv2
Judged at2026-09-23 11:41 UTC