index / choyiny/decidebench
decidebench
Benchmark of decision models (JEV and its open alternatives) and LLMs: accuracy, cost per task and latency on 400 contrastive decisions
Research and EvalsRoutingListedPython
Repository
Stars0
LanguagePython
LicenseMIT
Created2026-09-28
Last pushed2026-10-01
Officialno
Judgment
Genuine Jev project0.95
gate 0.5
Uses Jev at runtime0.49
Is a meta list0.09
Is a reimplementation0.20
Quality scales
Substance2.05 / 3
gate 0.5
Docs2.40 / 3
Novelty2.96 / 3
Composite score0.80
Category distribution
Research and Evals1.00
Integrations0.00
Other0.00
Games and Simulation0.00
Applications0.00
Agent and Dev Tooling0.00
SDKs and Clients0.00
Learning0.00
Assigned categoryResearch and Evals
Category probability1.00
Category confidence1.00
PatternRouting
Pattern confidence0.29
Curation
StatusListed
Reasongate passed
README pickno
Overridenone
Sources
- code:"jev-latest"
Provenance
Modeljev-1.13.0
Question setv2
Judged at2026-10-01 12:22 UTC