Jev technical blog and early access
The end of Diogo’s launch thread links to the official technical blog and early-access entry point for further reading.
Local System One toy: Jev-shaped API with Qwen 0.5B
The author shows a local System One toy using a Jev-shaped API and Qwen 0.5B, where both chess sides model only over legal moves; Stockfish 5k SFT raises move accuracy from 78% to 82%, with long multi-question prefix KV around 150ms.
Local System One toy (Jev-shaped API, Qwen 0.5B). Chess: both sides model-only over legal choices. Stockfish 5k SFT → move acc 78→82%. Prefix KV: long multi-Q ~150ms. https://github.com/fritzprix/systemone-lite…
· 0 likesOpen on X
More like this
Getting started with Jev: waitlist and Agent Skill
A community guide walks developers through requesting access and installing the TypeSafe Agent Skill.
Reverse-engineered a Jev-like architecture with open-source training repo
The author says they reverse-engineered a Jev-like architecture from its type and shares a GitHub repo for training your own jevlikes. Link
