# Human Ranking of Explanation Quality on VidGuard (20 videos)

Source: [Human Ranking of Explanation Quality on VidGuard (20 videos)](<https://www.sota2.com/research/sota/human-ranking-of-explanation-quality-on-vidguard-20-videos>)

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

Data updated: 2026-03-06T07:18:21.972Z

Evaluation of model explanation quality based on human ranking results collected from a specific 20-video subset of the VidGuard dataset. The primary metric is the average rank assigned by human evaluators.

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 | Human Ranking of Explanation Quality |
| Dataset (including any stated split) | VidGuard human evaluation subset (20 videos) |
| Metric | Average Rank |
| Unit | SCALAR |
| Direction | Lower 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

- [Average Rank](<https://www.sota2.com/research/sota/human-ranking-of-explanation-quality-on-vidguard-20-videos.md?metric=Average%20Rank>): Lower is better; unit: SCALAR.

## Recorded results

| Method | Average Rank | Result date | Report paper |
| --- | --- | --- | --- |
| VidGuard-R1 | 1.67 | Not recorded | 68fa7be71e209131cea8b388 |
| GPT-4o | 2.08 | Not recorded | 68fa7be71e209131cea8b388 |
| Qwen2.5-VL-72B | 2.22 | Not recorded | 68fa7be71e209131cea8b388 |

### VidGuard-R1

Average Rank: 1.67. Result ID: 69aa7f306dbb7075b953d013.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b388.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b388.md>)

Recorded evaluation configuration: {"training\_method":"GRPO"}. This record does not assert independent reproduction.

### GPT-4o

Average Rank: 2.08. Result ID: 69aa7f306dbb7075b953d014.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b388.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Qwen2.5-VL-72B

Average Rank: 2.22. Result ID: 69aa7f306dbb7075b953d015.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b388.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa773a1e209131cea71d3c.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

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