# Scientific Paper Feedback Generation on Human Evaluation Feedback Level

Source: [Scientific Paper Feedback Generation on Human Evaluation Feedback Level](<https://www.sota2.com/research/sota/scientific-paper-feedback-generation-on-human-evaluation-feedback-level>)

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

Data updated: 2026-04-15T07:42:16.107Z

Evaluation of model performance on generating feedback for scientific papers, measured by human assessments of Validity, Actionability, Specificity, and Helpfulness.

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 | Scientific Paper Feedback Generation |
| Dataset (including any stated split) | Human Evaluation Feedback Level |
| Metric | Validity Rate |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 3 of 3 |
| 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

- [Validity Rate](<https://www.sota2.com/research/sota/scientific-paper-feedback-generation-on-human-evaluation-feedback-level.md?metric=Validity%20Rate>): Higher is better; unit: PERCENT.

- [Actionability Rate](<https://www.sota2.com/research/sota/scientific-paper-feedback-generation-on-human-evaluation-feedback-level.md?metric=Actionability%20Rate>): Higher is better; unit: PERCENT.

- [Specificity Score](<https://www.sota2.com/research/sota/scientific-paper-feedback-generation-on-human-evaluation-feedback-level.md?metric=Specificity%20Score>): Higher is better; unit: SCALAR.

- [Helpfulness Score](<https://www.sota2.com/research/sota/scientific-paper-feedback-generation-on-human-evaluation-feedback-level.md?metric=Helpfulness%20Score>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Validity Rate | Result date | Report paper |
| --- | --- | --- | --- |
| Gemini-3-flash | 72.3 | Not recorded | 69df1b967d0172140ec7b901 |
| GOODPOINT-DPO | 58.1 | Not recorded | 69df1b967d0172140ec7b901 |
| Qwen3-8B | 41.5 | Not recorded | 69df1b967d0172140ec7b901 |

### Gemini-3-flash

Validity Rate: 72.3. Result ID: 69df3fea47f34ce924c7b638.

[Reporting paper](<https://www.sota2.com/research/paper/69df1b967d0172140ec7b901.md>)

Recorded evaluation configuration: {"n (per model)":"65"}. This record does not assert independent reproduction.

### GOODPOINT-DPO

Validity Rate: 58.1. Result ID: 69df3fea47f34ce924c7b639.

[Reporting paper](<https://www.sota2.com/research/paper/69df1b967d0172140ec7b901.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69df1b967d0172140ec7b901.md>)

Recorded evaluation configuration: {"n (per model)":"62"}. This record does not assert independent reproduction.

### Qwen3-8B

Validity Rate: 41.5. Result ID: 69df3fea47f34ce924c7b63a.

[Reporting paper](<https://www.sota2.com/research/paper/69df1b967d0172140ec7b901.md>)

Recorded evaluation configuration: {"n (per model)":"65"}. This record does not assert independent reproduction.

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