calibrated-probabilities

A curated list of tools built for Jev — TypeSafe AI's System One model for typed decisions.

🔥🔥 Awesome Jev: open-source models, projects, benchmarks, and independent evaluations for Jev and System One models.

Small decision models (jev style) for latency critical scenarios

离线可用的本地类型化决策：4 核 CPU 单题 15.6ms。Local & offline Jev / System One inference on CPU — ONNX + INT8, no torch at runtime. 支持 laya / kev / PlayJev

Wald-Q4B: open-weight 4B decision model. Calibrated probability for every option, Jev-compatible /v1/systemone API, self-hosted. Weights on Hugging Face.

Independent JEV-inspired System One-style typed decision research for clinical evidence, biomedical NLP, calibrated probabilities, and sleep signals

Find out which qualities of your writing actually predict engagement. Rates every post you have published against a pre-registered rubric using Jev's calibrated judgments, then tests those ratings against your real engagement numbers. Refuses to report findings your sample cannot support.

Hands-on System One decision lab: Jev-compatible Noul/Choice/Score over local open-weight Laya, with playground UI, routing/eval examples, and CI smoke tests.

Beveridge: System One decision models with open weights. Typed yes/no, choice and score decisions with calibrated probabilities, on your own machines. pip install bev-decisions

Playground for the Jev (TypeSafe System One) decision model via OpenRouter: typed questions (noul/choice/score), calibrated probabilities, latency/tokens/cost metrics, and Postman/cURL export.

Calibrated typed decisions from one forward pass of a language model

Open self-hosted System One decision API (Jev wire-compatible) + agent skill — 87.9% JevBench, beats Jev on accuracy and score