Jev Laya

A Python study collection for System One models: notes on Jev and its API primitives, a comparison against Laya, Cerebellum-2B, Kev-0.5B, BERT and XGBoost, and working demos including a local Laya run and an OpenJev API demo.

A minimal local dashboard and proxy for System One decision models: send one state with up to eight typed questions to Jev (TypeSafe API) or self-hosted Laya and get structured choice, score and noul answers rendered as charts.

A Chinese-language research collection on System One decision models: it compares JEV and Laya academically and operationally, maps enterprise adoption, teaches Laya locally from setup to fine-tuning, and reads the Laya source. Research finished 2026-09-24.

A self-hosted System One service that runs Laya through MLX on Apple silicon and keeps the model's native request and response format, exposing fast typed decisions to your own devices over a private Tailscale HTTPS endpoint.

A link collection rather than a codebase: babybear-labs gathers Jev architecture reading, Laya alternatives and ports, playgrounds, tiny decision-model projects and apps built on System One decision models, with no README prose beyond the links.

HarcoChen's VS Code client for DeepSeek Harness with native diffs, approval cards, persistent sessions, and a bundled Laya/Jev integration for loop guarding, done-gating, tool pruning, and typed decision tools.

Run Laya locally on Bun with native ONNX inference and Hugging Face tokenizers: an HTTP server implementing TypeSafe's decision API plus an English/Persian playground with 16 demos, usable through the official @typesafe-ai/sdk with a custom baseURL.

Benchmark comparing TypeSafe Jev 1.13 with the local Laya MLX typed-decision model across 1,470 synthetic items and eight tasks: accuracy, calibration, latency and throughput, with scripts, datasets, predictions and the full report in the repo.

A local-first Codex plugin and skill that gives AI conversations a daily brain budget, scored locally by Laya rather than Jev. Windows tray card, optional macOS menu bar or web dashboard, mood memes, and English and Chinese interfaces; Python 3.10+.

cklxx fine-tuned Laya into the decision head of browser-use/jev-ultrafast: one bidirectional encoder pass per step picks the browser operation and target element, and the server answers Jev's /v1/systemone requests identically.

Kitsuno's engineering write-up on training an open-weight Laya model to take over parts of its Jev usage: eight days of work, run in shadow beside Jev on every job-side request, without training on Jev's outputs.

A directory of 2,086 open-source projects built on Laya, the open System One decision model, with category and platform filters, star counts, a live laya-mlx demo, and a hosted Laya + Clef API with $1 free credit.

An independent Laya-versus-Jev comparison on one RTX 5090: phishing detection over 2,000 emails, the typed-decisions suite, latency scaling to 100 questions per pass and eleven probe assertions.

Browser Pac-Man race where a deterministic three-junction simulation filters fatal routes, then Jev makes one typed Choice among the surviving route IDs while the server rejects any answer outside the supplied set. Self-hosted Laya runs the same contract for comparison.