# Mathematical Reasoning on Geo3k Source: [Mathematical Reasoning on Geo3k]() Collection index: [llms.txt]() Data updated: 2026-06-09T09:43:58.476Z Evaluation of model performance on mathematical reasoning tasks using the Geo3k dataset, 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 | Mathematical Reasoning | | Dataset (including any stated split) | Geo3k | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 18 of 18 | | 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](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | NoisyRollout-7B | 52 | Not recorded | 6a279e6e90966ade919bfac6 | | PRPO | 46.9 | Not recorded | 6a279e6e90966ade919bfac6 | | VPPO | 44.4 | Not recorded | 6a279e6e90966ade919bfac6 | | PAPO-D | 44.1 | Not recorded | 6a279e6e90966ade919bfac6 | | R1-ShareVL-7B | 41.2 | Not recorded | 6a279e6e90966ade919bfac6 | | VL-Rethinker-7B | 40.7 | Not recorded | 6a279e6e90966ade919bfac6 | | MM-Eureka-7B | 40.3 | Not recorded | 6a279e6e90966ade919bfac6 | | PAPO-G | 40.3 | Not recorded | 6a279e6e90966ade919bfac6 | | GRPO | 40.2 | Not recorded | 6a279e6e90966ade919bfac6 | | PRPO | 37.3 | Not recorded | 6a279e6e90966ade919bfac6 | | Qwen2.5-VL-7B | 36.7 | Not recorded | 6a279e6e90966ade919bfac6 | | DAPO | 35.9 | Not recorded | 6a279e6e90966ade919bfac6 | | VPPO | 35.8 | Not recorded | 6a279e6e90966ade919bfac6 | | PAPO-D | 35.7 | Not recorded | 6a279e6e90966ade919bfac6 | | DAPO | 31.2 | Not recorded | 6a279e6e90966ade919bfac6 | | PAPO-G | 31.1 | Not recorded | 6a279e6e90966ade919bfac6 | | GRPO | 28.7 | Not recorded | 6a279e6e90966ade919bfac6 | | Qwen2.5-VL-3B | 18.8 | Not recorded | 6a279e6e90966ade919bfac6 | ### NoisyRollout-7B Accuracy: 52. Result ID: 6a27deeafbc373592c250738. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"7B","Rollout":"8"}. This record does not assert independent reproduction. ### PRPO Accuracy: 46.9. Result ID: 6a27deeafbc373592c250740. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"7B","Base Model":"Qwen2.5-VL-7B","Rollout":"8"}. This record does not assert independent reproduction. ### VPPO Accuracy: 44.4. Result ID: 6a27deeafbc373592c25073f. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Scale":"7B","Base Model":"Qwen2.5-VL-7B","Rollout":"8"}. This record does not assert independent reproduction. ### PAPO-D Accuracy: 44.1. Result ID: 6a27deeafbc373592c25073e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"7B","Base Model":"Qwen2.5-VL-7B","Rollout":"8"}. This record does not assert independent reproduction. ### R1-ShareVL-7B Accuracy: 41.2. Result ID: 6a27deeafbc373592c250739. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"7B","Rollout":"8"}. This record does not assert independent reproduction. ### VL-Rethinker-7B Accuracy: 40.7. Result ID: 6a27deeafbc373592c250737. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"7B","Rollout":"8"}. This record does not assert independent reproduction. ### MM-Eureka-7B Accuracy: 40.3. Result ID: 6a27deeafbc373592c250736. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"7B","Rollout":"8"}. This record does not assert independent reproduction. ### PAPO-G Accuracy: 40.3. Result ID: 6a27deeafbc373592c25073d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"7B","Base Model":"Qwen2.5-VL-7B","Rollout":"8"}. This record does not assert independent reproduction. ### GRPO Accuracy: 40.2. Result ID: 6a27deeafbc373592c25073b. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Scale":"7B","Base Model":"Qwen2.5-VL-7B","Rollout":"8"}. This record does not assert independent reproduction. ### PRPO Accuracy: 37.3. Result ID: 6a27deeafbc373592c250735. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"3B","Base Model":"Qwen2.5-VL-3B","Rollout":"8"}. This record does not assert independent reproduction. ### Qwen2.5-VL-7B Accuracy: 36.7. Result ID: 6a27deeafbc373592c25073a. [Reporting paper]() Recorded evaluation configuration: {"Model Scale":"7B","Rollout":"8"}. This record does not assert independent reproduction. ### DAPO Accuracy: 35.9. Result ID: 6a27deeafbc373592c25073c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"7B","Base Model":"Qwen2.5-VL-7B","Rollout":"8"}. This record does not assert independent reproduction. ### VPPO Accuracy: 35.8. Result ID: 6a27deeafbc373592c250734. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Scale":"3B","Base Model":"Qwen2.5-VL-3B","Rollout":"8"}. This record does not assert independent reproduction. ### PAPO-D Accuracy: 35.7. Result ID: 6a27deeafbc373592c250733. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"3B","Base Model":"Qwen2.5-VL-3B","Rollout":"8"}. This record does not assert independent reproduction. ### DAPO Accuracy: 31.2. Result ID: 6a27deeafbc373592c250731. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"3B","Base Model":"Qwen2.5-VL-3B","Rollout":"8"}. This record does not assert independent reproduction. ### PAPO-G Accuracy: 31.1. Result ID: 6a27deeafbc373592c250732. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"3B","Base Model":"Qwen2.5-VL-3B","Rollout":"8"}. This record does not assert independent reproduction. ### GRPO Accuracy: 28.7. Result ID: 6a27deeafbc373592c250730. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Scale":"3B","Base Model":"Qwen2.5-VL-3B","Rollout":"8"}. This record does not assert independent reproduction. ### Qwen2.5-VL-3B Accuracy: 18.8. Result ID: 6a27deeafbc373592c25072f. [Reporting paper]() Recorded evaluation configuration: {"Model Scale":"3B","Rollout":"8"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.