# Octalab Inc: jqv

> Octalab's open serving engine that reads an unmodified Qwen3-32B as a decision model: the state is prefilled once, each question runs as an isolated branch, and choice, yes/no and score answers come from the option-letter probabilities in one pass, with no generated text.

- Page: https://systemonemodels.ai/octalab/jqv
- API: https://api.systemonemodels.ai/v1/models/octalab/jqv
- Download: `pip install systemonemodels && systemone pull octalab/jqv`

## Facts

| | |
|---|---|
| Maker | Octalab Inc (https://systemonemodels.ai/octalab) |
| Decides | choice, score, noul |
| Architecture | jqv |
| Base model | qwen/qwen3-32b |
| Parameters | 32B |
| Licence | apache-2.0 |
| Availability | Open weights |
| Released | 2026-09-21 |
| Latest version | 0.1.0 |

## Reported evaluation

Suite: MMLU test, 800 items (Octalab's own evaluation: zero-shot packed engine on Qwen3-32B, one temperature fitted on 400 MMLU validation items). Numbers are the publisher's own.

- Decision accuracy: 80.9%
- Calibration error (ECE): 0.023

## Model card

<!-- generated by scripts/seed_catalog.py; edit content/models/catalog.yaml -->

# jqv (Qwen3-32B)

Octalab's open serving engine that reads an unmodified Qwen3-32B as a decision model: the state is prefilled once, each question runs as an isolated branch, and choice, yes/no and score answers come from the option-letter probabilities in one pass, with no generated text.

jqv adds no weights in its default configuration: it runs Qwen/Qwen3-32B (Apache-2.0) unchanged, so there are no files to pull here. One temperature fitted on 400 MMLU validation items scales the probabilities; the maker reports that this transfers only partly to harder decision items and hurts on a reading task. With JevBench's harness on the public items the maker measured easy 1.000, standard 0.958 and hard 0.622 (hard-tier ECE 0.107). An optional trained readout (a 40M-parameter LoRA plus a slot head, 148 MB) is a GitHub release, not on Hugging Face; with it the maker measures 82/111 on the public hard tier, but says its synthetic training data was designed after looking at that tier's misses. The server speaks TypeSafe's /v1/systemone wire format; there is no hosted API. The project calls itself a proof-of-concept reconstruction of Jev's inference design. Third-party: JevBench v1.4.2.2 ranks the zero-shot configuration 19th of 91 (score 44.35; 80.1% public, 28.2% sealed). By Hajime Imura, published under the Octalab Inc GitHub organisation; held on 30 September as a borderline shim and listed now on the AnyJev and Cygnet precedent.

## What it decides

- **choice** — picks one option from a set
- **score** — places the input on an ordered scale
- **noul** — answers a yes/no question with one probability

## At a glance

| | |
|---|---|
| Parameters | 32B |
| Base model | `qwen/qwen3-32b` |
| Maker | Octalab Inc |
| Released | 2026-09-21 |
| License | apache-2.0 |
| Reported accuracy | 80.9% |
| Reported latency | 0.65 s p50 per request on an Apple M5 Max (Qwen3-32B BF16, PyTorch/MPS), JevBench public tiers, one request at a time |

## Read more

- [Code and results](https://github.com/Octalab-Inc/jqv)
- [Full report](https://github.com/Octalab-Inc/jqv/blob/main/docs/report.md)
- [Trained readout (targeted-v2-32b release)](https://github.com/Octalab-Inc/jqv/releases/tag/targeted-v2-32b)
- [Qwen3-32B on Hugging Face](https://huggingface.co/Qwen/Qwen3-32B)

---

*This page was opened by System One for Octalab Inc, who can claim the organisation and take it over at any time.*

---

From System One Models — https://systemonemodels.ai/ · every System One model: https://systemonemodels.ai/system-one-models