# Text Quality on GPT-4o Prompt Set

Source: [Text Quality on GPT-4o Prompt Set](<https://www.sota2.com/research/sota/text-quality-on-gpt-4o-prompt-set>)

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

Data updated: 2026-06-01T08:47:32.274Z

Evaluation of model performance on the GPT-4o Prompt Set, measuring overall text quality using a percentage score.

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 | Text Quality |
| Dataset (including any stated split) | GPT-4o Prompt Set |
| Metric | Overall Quality Score |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 2 of 2 |
| 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

- [Overall Quality Score](<https://www.sota2.com/research/sota/text-quality-on-gpt-4o-prompt-set.md?metric=Overall%20Quality%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Overall Quality Score | Result date | Report paper |
| --- | --- | --- | --- |
| Persona-based evaluation framework | 85.6 | Not recorded | 6a1d11eadbd7150506f83e4b |
| Human Consensus Baseline | 85 | Not recorded | 6a1d11eadbd7150506f83e4b |

### Persona-based evaluation framework

Overall Quality Score: 85.6. Result ID: 6a1d4711fbc373592c168fab.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d11eadbd7150506f83e4b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a1d11eadbd7150506f83e4b.md>)

Recorded evaluation configuration: {"Evaluation Source":"Persona","Source Description":"Persona-based creativity alignment"}. This record does not assert independent reproduction.

### Human Consensus Baseline

Overall Quality Score: 85. Result ID: 6a1d4711fbc373592c168faa.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d11eadbd7150506f83e4b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa765b1e209131cea6c9ae.md>)

Recorded evaluation configuration: {"Evaluation Source":"Baseline","Source Description":"Human-rated creativity consensus"}. This record does not assert independent reproduction.

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