# Subjective Rubric-based Scoring on OmniScore overall (test)

Source: [Subjective Rubric-based Scoring on OmniScore overall (test)](<https://www.sota2.com/research/sota/subjective-rubric-based-scoring-on-omniscore-overall-test>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-04-08T07:49:17.537Z

Evaluation of model performance using subjective rubric-based scoring criteria on the OmniScore overall test set, including accuracy metrics (MAE, L-MAE, Acc@±1) and efficiency metrics (Time, Cost).

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Subjective Rubric-based Scoring |
| Dataset (including any stated split) | OmniScore overall (test) |
| Metric | MAE |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 5 of 5 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [MAE](<https://www.sota2.com/research/sota/subjective-rubric-based-scoring-on-omniscore-overall-test.md?metric=MAE>): Lower is better; unit: SCALAR.

- [L-MAE](<https://www.sota2.com/research/sota/subjective-rubric-based-scoring-on-omniscore-overall-test.md?metric=L-MAE>): Lower is better; unit: SCALAR.

- [Acc@ ±1](<https://www.sota2.com/research/sota/subjective-rubric-based-scoring-on-omniscore-overall-test.md?metric=Acc%40%20%C2%B11>): Higher is better; unit: PERCENT.

- [Time](<https://www.sota2.com/research/sota/subjective-rubric-based-scoring-on-omniscore-overall-test.md?metric=Time>): Lower is better; unit: TIME.

- [Cost](<https://www.sota2.com/research/sota/subjective-rubric-based-scoring-on-omniscore-overall-test.md?metric=Cost>): Lower is better; unit: SCALAR.

## Recorded results

| Method | MAE | Result date | Report paper |
| --- | --- | --- | --- |
| mxbai-large | 0.78 | Not recorded | 69d5e1237d0172140ec75b38 |
| DeBERTa-v3 | 0.79 | Not recorded | 69d5e1237d0172140ec75b38 |
| Ettin-32M | 0.8 | Not recorded | 69d5e1237d0172140ec75b38 |
| Ettin-17M | 0.84 | Not recorded | 69d5e1237d0172140ec75b38 |
| Gemini-3.1-Flash-Lite | 1.01 | Not recorded | 69d5e1237d0172140ec75b38 |

### mxbai-large

MAE: 0.78. Result ID: 69d6087447f34ce924be5592.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1237d0172140ec75b38.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1237d0172140ec75b38.md>)

Recorded evaluation configuration: {"Params":"335M","Ctx":"1K"}. This record does not assert independent reproduction.

### DeBERTa-v3

MAE: 0.79. Result ID: 69d6087447f34ce924be5593.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1237d0172140ec75b38.md>)

Recorded evaluation configuration: {"Params":"86M","Ctx":"512"}. This record does not assert independent reproduction.

### Ettin-32M

MAE: 0.8. Result ID: 69d6087447f34ce924be5595.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1237d0172140ec75b38.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1237d0172140ec75b38.md>)

Recorded evaluation configuration: {"Params":"32M","Ctx":"8K"}. This record does not assert independent reproduction.

### Ettin-17M

MAE: 0.84. Result ID: 69d6087447f34ce924be5594.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1237d0172140ec75b38.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1237d0172140ec75b38.md>)

Recorded evaluation configuration: {"Params":"17M","Ctx":"8K"}. This record does not assert independent reproduction.

### Gemini-3.1-Flash-Lite

MAE: 1.01. Result ID: 69d6087447f34ce924be5596.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1237d0172140ec75b38.md>)

Recorded evaluation configuration: {"Ctx":"65k"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.