App · Benchmarks & research
One judge call vs twelve dimension scores
One direct Jev question per row against 12–14 Jev-scored dimensions with locally fitted weights on three classification tasks: 5,477 test rows,…
Open agentjournal.dev ↗How builders describe it
One direct Jev question per row against 12–14 Jev-scored dimensions with locally fitted weights on three classification tasks: 5,477 test rows, 25,174 Jev calls, $1.43. Decomposition wins on Japanese NLI (0.9076 vs 0.8373) but flags about 25× more hard benign rows as attacks (37.2% vs 1.5%).
The decision Jev makes
Benchmark questions with known answers, to check accuracy and confidence.
Where it fits
Head-to-head tests, calibration studies and independent research into how well Jev decides, how fast, and at what cost. All 332 benchmarks & research projects →
Related projects
Benchmarks & researchJev 1.13.0 against Hopper on Benchmark HeavenHopper wins three columns, Jev keeps Intelligence.Site · Benchmark Heaven
Benchmarks & researchValenTrain a Jev-like multimodal model by yourself.GitHub · ★ 51 · Liuziyu77
Benchmarks & researchjev-dimabsaTypeSafe Jev baseline for DimABSA (SemEval-2026 Task 3) subtask 1: zero-shot and 3-shot valence-arousal regression.GitHub · ★ 6 · ZhangYiqun018
Benchmarks & researchreflexbenchReflexBench — open benchmark and evaluation harness for System One models and typed decision engines.GitHub · ★ 4 · brida-ai
Benchmarks & researchjev-guardrailsResearch/eval repo that runs the same mock carrier support agent and 25 behavioural guardrail rules behind two backends—chat-model JSON judge vs…GitHub · ★ 3 · deepansh-saxena
Benchmarks & researchtypesafe-jev-calibrate-for-code-reviewAbout calibrating Jev for code reviews.GitHub · ★ 3 · Selmar