Lineage & credits
- AutoTrust AI — Blocks-of-Experts recipe, the System-1 decision LoRA (r16/r32) + 24-slot decision head, and the serve_decide.py decision harness, from autotrust/JEV-27B and autotrust/JEV-27B-VL (Apache-2.0).
- huihui-ai — the abliterated (uncensored) Qwen3.8-27B backbone used here (Huihui-Qwen3.8-27B-abliterated, Apache-2.0; layers 18-51 ablated, vision tower untouched).
- SargeDev — jev-distill-corpus-v3, the 740,957-row calibrated typed-decision corpus AutoTrust's JEV models were trained on (Apache-2.0). The quantization calibration prompts additionally sampled 384 prompts from SargeDev/solar-decisions-corpus-v4 (text only).
- Qwen — Qwen3.8-27B base (Apache-2.0).
- TypeSafe AI — Jev 1.13, the original closed teacher behind the System-1 typed-decisions framing (referenced; not redistributed).
What this is
NVFP4 (compressed-tensors, W4A4 + FP8 down_proj) of the vision flagship — the VL composite
(vision tower + merger present, MTP dropped per AutoTrust's convention) with the AutoTrust
System-1 decision block merged into huihui's abliterated Qwen3.8-27B backbone. The text tower is
quantized (NVFP4 attention/MLP/GDN projections + FP8 dynamic down_proj); the vision tower ships
unquantized bf16 (model-visual-bf16.safetensors) so image fidelity is untouched.
The decision adapter is already merged into these weights — never mount any additional decision
adapter on top. The decision machinery ships in the parent text repo
SargeDev/JEV-27B-Uncensored (head.safetensors,
decision_head.json, judge_config.json, calibration.json, calibration_mergedfit.json,
serve_decide.py, read_decision.py); the head projects the same 5120-dim text-tower hidden state.
A quant-specific calibration battery for NVFP4 precision is pending — treat the refit temperatures as close but not yet validated at this precision. The table below was measured on the bf16 weights before quantization. Vision decision battery results will be appended when finalized.
Honest evaluation (the backbone-swap trade, measured)
Measured on the held-out test_set_30k (27,695 rows, D1-excluded) of jev-distill-corpus-v3,
with AutoTrust's own acceptance floors as the yardstick (bf16 weights, pre-quantization; clean
bare-weights 30k gate, 2026-10-10 — earlier published numbers came from a mis-mounted adapter
configuration and have been superseded):
| metric | their JEV-27B (pristine Qwen) | this model (uncensored backbone) | their floor | status |
|---|---|---|---|---|
| noul AUROC | 0.9961 | 0.992 | >= 0.95 | PASS |
| noul top-1 | 0.962 | 0.948 | — | -1.4 pts |
| choice top-1 | 0.904 | 0.870 | >= 0.90 | FAIL |
| score top-1 | 0.890 | 0.814 | >= 0.891 | FAIL |
| overall KL | 0.019 | 0.040 | <= 0.15 | PASS |
| ECE (raw) | 0.0011 | 0.009 | <= 0.03 | PASS |
| ECE (refit T: 0.9/1.0/1.2) | — | noul 0.0022 / choice 0.0049 / score 0.0050 | <= 0.03 | ALL PASS |
noul (yes/no) judgment is essentially intact. The choice/score top-1 dips are the real cost of the abliteration backbone-swap (the residual stream the adapter was calibrated against is rewired); temperature refit recovers all calibration (ECE) floors. Uncensored behavior preserved: refusal rate 0.0 on a 10-prompt battery, base vs +JEV identical.
Training dataset
Trained on SargeDev/jev-distill-corpus-v3 — 740,957 rows of typed calibrated decisions (noul/choice/score) distilled from Jev 1.13 (System One). Fine-tuned on AutoTrust's JEV-27B System-1 decision block, merged into huihui-ai's abliterated Qwen3.8-27B backbone. Quantization calibration sampled 384 prompts from SargeDev/solar-decisions-corpus-v4 (text only).
Not for high-stakes decisions. Use confidence gating; route low-confidence calls to a stronger model or a human (AutoTrust's own caveat, still true here).
Serving
vllm serve . --quantization compressed-tensors --served-model-name SargeDev/JEV-27B-VL-Uncensored-NVFP4
System 2 (chat/code/vision) runs through the normal lm_head with the bundled chat template. For the
calibrated System-1 decision engine use the parent repo's read_decision.py recipe (bare-text
prompt -> last-position final hidden state -> head projection -> per-kind temperature).