# Multimodal Math Reasoning on R1-OV Bench

Source: [Multimodal Math Reasoning on R1-OV Bench](<https://www.sota2.com/research/sota/multimodal-math-reasoning-on-r1-ov-bench>)

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

Data updated: 2026-03-31T09:10:41.693Z

Evaluation of model performance on multimodal mathematical reasoning tasks using the R1-OV Bench dataset, typically measured by 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 | Multimodal Math Reasoning |
| Dataset (including any stated split) | R1-OV Bench |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 8 of 8 |
| 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

- [Accuracy](<https://www.sota2.com/research/sota/multimodal-math-reasoning-on-r1-ov-bench.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| AVATAR | 38.9 | Not recorded | 68fa7a631e209131cea82cee |
| AVATAR | 38.5 | Not recorded | 68fa7a631e209131cea82cee |
| VLAAThinker-VL-7B | 38.4 | Not recorded | 68fa7a631e209131cea82cee |
| AVATAR | 37.5 | Not recorded | 68fa7a631e209131cea82cee |
| AVATAR | 37 | Not recorded | 68fa7a631e209131cea82cee |
| Qwen2.5-VL-7B + GRPO | 36.8 | Not recorded | 68fa7a631e209131cea82cee |
| Qwen2.5-VL-7B | 34.9 | Not recorded | 68fa7a631e209131cea82cee |
| OpenVLThinker-7B | 34.7 | Not recorded | 68fa7a631e209131cea82cee |

### AVATAR

Accuracy: 38.9. Result ID: 69cb8eb247f34ce924b411ec.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a631e209131cea82cee.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a631e209131cea82cee.md>)

Recorded evaluation configuration: {"backbone":"Qwen2.5-VL-7B"}. This record does not assert independent reproduction.

### AVATAR

Accuracy: 38.5. Result ID: 69cb8eb247f34ce924b411ed.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a631e209131cea82cee.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a631e209131cea82cee.md>)

Recorded evaluation configuration: {"backbone":"Qwen2.5-VL-7B","ablation":"w/o Replay Buffer (on-policy)"}. This record does not assert independent reproduction.

### VLAAThinker-VL-7B

Accuracy: 38.4. Result ID: 69cb8eb247f34ce924b411e9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a631e209131cea82cee.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78051e209131cea765df.md>)

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

### AVATAR

Accuracy: 37.5. Result ID: 69cb8eb247f34ce924b411ee.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a631e209131cea82cee.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a631e209131cea82cee.md>)

Recorded evaluation configuration: {"backbone":"Qwen2.5-VL-7B","ablation":"w/o TAS (uniform credit)"}. This record does not assert independent reproduction.

### AVATAR

Accuracy: 37. Result ID: 69cb8eb247f34ce924b411ef.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a631e209131cea82cee.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a631e209131cea82cee.md>)

Recorded evaluation configuration: {"backbone":"Qwen2.5-VL-7B","ablation":"w/o Hinting"}. This record does not assert independent reproduction.

### Qwen2.5-VL-7B + GRPO

Accuracy: 36.8. Result ID: 69cb8eb247f34ce924b411eb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a631e209131cea82cee.md>)

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

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

### Qwen2.5-VL-7B

Accuracy: 34.9. Result ID: 69cb8eb247f34ce924b411ea.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a631e209131cea82cee.md>)

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

Recorded evaluation configuration: {"status":"Baseline"}. This record does not assert independent reproduction.

### OpenVLThinker-7B

Accuracy: 34.7. Result ID: 69cb8eb247f34ce924b411e8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a631e209131cea82cee.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a251e209131cea814a7.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.
