# Multimedia Audio Reasoning on MMAR Single Model Track

Source: [Multimedia Audio Reasoning on MMAR Single Model Track](<https://www.sota2.com/research/sota/multimedia-audio-reasoning-on-mmar-single-model-track>)

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

Data updated: 2026-06-09T08:33:34.486Z

Evaluation of Multimedia Audio Reasoning performance on the MMAR Single Model Track, reporting results based on Rubrics and Accuracy.

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 | Multimedia Audio Reasoning |
| Dataset (including any stated split) | MMAR Single Model Track |
| Metric | Rubrics Score |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 4 of 4 |
| 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

- [Rubrics Score](<https://www.sota2.com/research/sota/multimedia-audio-reasoning-on-mmar-single-model-track.md?metric=Rubrics%20Score>): Higher is better; unit: PERCENT.

- [Accuracy](<https://www.sota2.com/research/sota/multimedia-audio-reasoning-on-mmar-single-model-track.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Rubrics Score | Result date | Report paper |
| --- | --- | --- | --- |
| Qwen3-Omni + RL (GRPO) | 65.29 | Not recorded | 6a279e1a90966ade919bf6fe |
| Qwen3-Omni + Attn. Manip. | 62.55 | Not recorded | 6a279e1a90966ade919bf6fe |
| Qwen3-Omni + LoRA | 62.22 | Not recorded | 6a279e1a90966ade919bf6fe |
| TinyGiantALM | 23.77 | Not recorded | 6a279e1a90966ade919bf6fe |

### Qwen3-Omni + RL (GRPO)

Rubrics Score: 65.29. Result ID: 6a27cf0cfbc373592c249033.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e1a90966ade919bf6fe.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b881e209131cea88f3d.md>)

Recorded evaluation configuration: {"Size (Est.)":"\>30B","Evaluation Protocol":"Single Model Track"}. This record does not assert independent reproduction.

### Qwen3-Omni + Attn. Manip.

Rubrics Score: 62.55. Result ID: 6a27cf0cfbc373592c249034.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e1a90966ade919bf6fe.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b881e209131cea88f3d.md>)

Recorded evaluation configuration: {"Size (Est.)":"\>30B","Evaluation Protocol":"Single Model Track"}. This record does not assert independent reproduction.

### Qwen3-Omni + LoRA

Rubrics Score: 62.22. Result ID: 6a27cf0cfbc373592c249035.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e1a90966ade919bf6fe.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b881e209131cea88f3d.md>)

Recorded evaluation configuration: {"Size (Est.)":"\>30B","Evaluation Protocol":"Single Model Track"}. This record does not assert independent reproduction.

### TinyGiantALM

Rubrics Score: 23.77. Result ID: 6a27cf0cfbc373592c249036.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e1a90966ade919bf6fe.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a279e1a90966ade919bf6fe.md>)

Recorded evaluation configuration: {"Size (Est.)":"1.5B","Evaluation Protocol":"Single Model Track"}. This record does not assert independent reproduction.

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