jevify-qwen3.5-4b

A System One decision model: it does not write text. It reads a state, answers typed questions, and returns calibrated probability distributions your code can branch on.

primitive	question	answer
choice	which of these K options?	probabilities over the options + confidence
score	where on these K ordered levels?	probabilities over levels, expected score, confidence
noul	is this true?	a single P(yes)

This repo holds only what Jevify adds to Qwen/Qwen3.5-4B: 2,891,275 parameters of trained decision heads (11.6 MB) plus the calibration recipe. The backbone is pulled from its own repo at load time, so nothing is duplicated or relicensed.

Results on jev-bench

Scored on all 22,773 test records, against TypeSafe's Jev 1.13.0 on the identical records.

model	held-out sources: acc / ECE	sources seen in training: acc / ECE
Qwen3.5-4B (Tier 1 residual) — Tier 1 heads	0.749 / 0.107	0.680 / 0.082
Jev 1.13.0 — API (zero-shot)	0.835 / 0.090	0.694 / 0.122
Qwen/Qwen3.5-4B — Tier 0 (recipe refit w/o held-out)	0.714 / 0.139	0.641 / 0.087

The held-out row above is the published seed. Five seeds of this configuration give held-out accuracy 0.715 ± 0.025, level with Tier 0 (0.714), so the held-out gain shown here is this seed's, not the method's. The trained-source gain is robust: 0.688 ± 0.005 against Tier 0's 0.641, every seed. FINDINGS 6.3b

Across the whole benchmark: macro accuracy 0.698 against Jev's 0.733, ECE 0.089 against 0.113, and 0.360 mean distance to human label distributions against Jev's 0.432 — lower is better, and that last number is the one a calibration claim rests on.

Use it

pip install git+https://github.com/uspraveen/Jevify

from jevify import load_jevified

model = load_jevified("Praveenrajus/jevify-qwen3.5-4b")
answer = model.ask(
    state={"ticket": "I was charged twice for order A-104, please refund the duplicate."},
    questions={
        "dept": {"type": "choice", "instructions": "Which team should handle `ticket`?",
                 "criteria": {"billing": "Payments and refunds", "shipping": "Delivery problems", "other": None}},
        "refund": {"type": "noul", "instructions": "Does `ticket` ask for a refund?"},
        "anger": {"type": "score", "instructions": "How angry is the customer?",
                  "criteria": ["calm", "annoyed", "furious"]},
    },
)
print(answer["dept"]["choice"], answer["dept"]["confidence"])
print(answer["refund"]["noul"])

As a drop-in for the TypeSafe API

jevify-serve --model Praveenrajus/jevify-qwen3.5-4b --port 8000

TYPESAFE_BASE_URL=http://localhost:8000 python your_existing_typesafe_code.py

The official typesafe-sdk works against this unchanged — that is a test in the repo.

How it was built

Decision heads read the backbone's hidden state at each option's own line, so they score what an option means rather than how likely its identifier token is. They are applied as a residual on the model's own log-score — score_i = w·lm_i + f(...) with f zero-initialized — so training starts exactly at the untrained baseline and can only add to it. Trained on 8,685 records from 16 sources with at most 16 options each, which is what keeps the head usable at any K.

Training optimizes strictly proper scoring rules directly — log score for Choice and Noul, ranked probability score for the ordinal Score — so calibration is the objective rather than a post-hoc repair. No reinforcement learning is involved: with a differentiable head the calibration objective is just a loss.

Six sources were held out of training entirely so generalization to unseen question types is measured rather than assumed. Full method, findings and limitations: github.com/uspraveen/Jevify · FINDINGS.md

Limitations

• English-first, text only, following the benchmark it was tuned on.
• Ordinal (score) questions on scales unlike those in training are the weakest case.
• The heads are trained on jev-bench's own train splits, so "held out" means held-out source, not a wholly different data universe.
• Seed variance is real and measured: see the paragraph under the results table. Differences of a few hundredths between models are within it.