# Kotoba Labs: open-jev-deberta-v3-large

> An independent Jev-shaped reproduction on DeBERTa-v3-large. One state and any number of choice, score and yes/no questions go in, and a calibrated distribution per question comes out of one forward pass. Public gold labels only.

- Page: https://systemonemodels.tech/kotoba-labs/open-jev-deberta-v3-large
- API: https://api.systemonemodels.tech/v1/models/kotoba-labs/open-jev-deberta-v3-large
- Download: `pip install systemonemodels && systemone pull kotoba-labs/open-jev-deberta-v3-large`

## Facts

| | |
|---|---|
| Maker | Kotoba Labs (https://systemonemodels.tech/kotoba-labs) |
| Decides | choice, score, noul, classify |
| Architecture | open-jev |
| Base model | microsoft/deberta-v3-large |
| Parameters | 434M |
| Context | 512 tokens |
| Licence | apache-2.0 |
| Availability | Open weights |
| Released | 2026-09-18 |
| Latest version | 2026.09 |

## Reported evaluation

Suite: Kotoba held-out test, seen question types (banking77, SST-5, BoolQ). Numbers are the publisher's own.

- Decision accuracy: 85.4%
- Calibration error (ECE): 0.022
- Median latency: 28 ms

## Model card

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

# open-jev-deberta-v3-large

An independent Jev-shaped reproduction on DeBERTa-v3-large. One state and any number of choice, score and yes/no questions go in, and a calibrated distribution per question comes out of one forward pass. Public gold labels only.

Trained for one epoch on banking77, SST-5 and BoolQ (42,000 questions), it scores 85.4% with a calibration error of 0.022 on held-out states for question types it has seen, and 69.0% (calibration error 0.035) on 4,012 questions with new instructions and option sets. Context is 512 tokens, with the state cut to 256; choice takes up to 255 options and score 2 to 10 levels. Ten questions on one state take 28 ms end to end on an H100; four take 1.8 s on an M1 Max CPU. The authors note its numbers are measured on public gold labels and are not comparable to Jev's.

## 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 calibrated probability
- **classify** — assigns a category from a fixed taxonomy

## At a glance

| | |
|---|---|
| Parameters | 434M |
| Base model | `microsoft/deberta-v3-large` |
| Maker | Kotoba Labs |
| Released | 2026-09-18 |
| License | apache-2.0 |
| Reported accuracy | 85.4% |
| Reported latency | 28 ms for ten questions on an H100; 1.8 s for four on an M1 Max CPU |

## Get the weights

```bash
pip install systemonemodels
systemone pull kotoba-labs/open-jev-deberta-v3-large
```

The files are served from the maker's Hugging Face repository, [`com-kotobalabs/open-jev-deberta-v3-large`](https://huggingface.co/com-kotobalabs/open-jev-deberta-v3-large), and verified against the checksums recorded here.

## Read more

- [Source and training code](https://github.com/kotoba-lang/typed-decisions)
- [Weights on Hugging Face](https://huggingface.co/com-kotobalabs/open-jev-deberta-v3-large)

---

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

---

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