index / TrustifAI/typed_evals
typed_evals
Fast, typed Python framework for Jev-as-a-Judge/Decision-model evaluation for LLMs, RAG, and AI agents along with human-label calibration.
Research and Evals?Feature ScoringListedPython
Repository
Stars16
LanguagePython
LicenseMIT
Created2026-09-20
Last pushed2026-10-29
Officialno
Judgment
Genuine Jev project0.87
gate 0.5
Uses Jev at runtime0.78
Is a meta list0.03
Is a reimplementation0.07
Quality scales
Substance2.72 / 3
gate 0.5
Docs2.13 / 3
Novelty1.78 / 3
Composite score0.81
Category distribution
Jev's confidence in this category is below 0.5, so the category is shown with a question mark. It is not a listing condition: this entry's status was decided by genuine and substance alone.
Research and Evals0.54
SDKs and Clients0.28
Agent and Dev Tooling0.11
Integrations0.04
Applications0.01
Other0.01
Games and Simulation0.00
Learning0.00
Assigned categoryResearch and Evals (uncertain)
Category probability0.54
Category confidence0.47
PatternFeature Scoring
Pattern confidence0.44
Curation
StatusListed
Reasongate passed
README pickno
Overridenone
Sources
- search:topic:jev
- search:jev in:name,description,readme created:2026-09-20
Provenance
Modeljev-1.14.0
Question setv2
Judged at2026-10-29 12:17 UTC