jev-lm (uhhfeef)

A character-level language model built on Jev, a System One classifier

How it uses Jev

Jev is a System One model from TypeSafe: it picks one option from a defined set and returns a calibrated probability over it — the same shape as a character-level softmax. So the vocabulary becomes 28 Choice options ( a-z , space, stopping) and generation is the ordinary autoregressive loop.

Source: community catalog. Details and metrics reflect the published community snapshot and are not a performance endorsement. Documentation details extracted from GitHub · the README.

Keep exploring

Similar to jev-lm (uhhfeef)

Browse all
simple-jev project preview
Applications
Turn any open model into a classifier/jev endpoint
jev-mail-classifier project preview
Classify your inbox with Jev (TypeSafe's System One model) — tag, move, flag, and notify, all config-driven.
LLM2Jev project preview
Applications
Adapt local language models into Jev-compatible structured decision engines with Choice, Score, and Noul outputs powered by prefill-only binary inference.
jev-freeform project preview
Benchmarks & research
An observable raw-character chat experiment powered entirely by TypeSafe Jev Choice
jev-bot project preview
Benchmarks & research
Self-hosted Jev decision workbench and Feishu bot: automatic choices, probabilities, and experimental word/character writing.
jev-lm project preview
Benchmarks & research
A word-level language model whose output layer is Jev: n-gram drafter, Noul chunk verification, bits-per-token eval
pi-typesafe project preview
Agent tooling
Pi coding-agent extension built on the TypeSafe AI System One API (Jev)
JEV-language project preview
Demos & playgrounds
Give your Jev language
agi-jev-containment project preview
Games & simulations
AGI JEV Detection — local AI agent monitor: chain-level malicious-agent detection (TypeSafe Jev + Sentinel), escalate-only L1–L5 containment, Neo4j forensics, AngryRobot dashboard. HackSpain 2026.