# Geometry problem solving on Geomverse (BoN@3 accuracy) Source: [Geometry problem solving on Geomverse (BoN@3 accuracy)]() Collection index: [llms.txt]() Data updated: 2026-04-21T12:56:27.471Z Evaluation of model performance on the Geomverse dataset for geometry problem solving tasks, specifically measured using BoN@3 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 | Geometry problem solving | | Dataset (including any stated split) | Geomverse | | Metric | BoN@3 Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 17 of 17 | | 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 - [BoN@3 Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | BoN@3 Accuracy | Result date | Report paper | | --- | --- | --- | --- | | GeometryZero-7B | 18.23 | Not recorded | 68fa79d01e209131cea7f82e | | Qwen2.5-7B-Instruct + GRPO | 16.93 | Not recorded | 68fa79d01e209131cea7f82e | | Qwen2.5-7B-Instruct + SFT | 15.36 | Not recorded | 68fa79d01e209131cea7f82e | | Qwen2.5-7B-Instruct | 14.76 | Not recorded | 68fa79d01e209131cea7f82e | | Qwen2.5-7B-Instruct + ToRL | 12.56 | Not recorded | 68fa79d01e209131cea7f82e | | Qwen2.5-3B-Instruct + GRPO | 12.13 | Not recorded | 68fa79d01e209131cea7f82e | | Qwen2.5-3B-Instruct + ToRL | 11.63 | Not recorded | 68fa79d01e209131cea7f82e | | GeometryZero-3B | 11.3 | Not recorded | 68fa79d01e209131cea7f82e | | Qwen2.5-3B-Instruct | 10.53 | Not recorded | 68fa79d01e209131cea7f82e | | Qwen2.5-3B-Instruct + SFT | 10.2 | Not recorded | 68fa79d01e209131cea7f82e | | GeometryZero-1.5B | 6.95 | Not recorded | 68fa79d01e209131cea7f82e | | G-Llava-7B | 6.23 | Not recorded | 68fa79d01e209131cea7f82e | | Qwen2.5-1.5B-Instruct + GRPO | 5.76 | Not recorded | 68fa79d01e209131cea7f82e | | Qwen2.5-1.5B-Instruct + ToRL | 5.26 | Not recorded | 68fa79d01e209131cea7f82e | | GNS-Llava-1.5-7B | 5.21 | Not recorded | 68fa79d01e209131cea7f82e | | Qwen2.5-1.5B-Instruct + SFT | 4.8 | Not recorded | 68fa79d01e209131cea7f82e | | Qwen2.5-1.5B-Instruct | 4.2 | Not recorded | 68fa79d01e209131cea7f82e | ### GeometryZero-7B BoN@3 Accuracy: 18.23. Result ID: 69e772de47f34ce924cf31e9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"7B"}. This record does not assert independent reproduction. ### Qwen2.5-7B-Instruct + GRPO BoN@3 Accuracy: 16.93. Result ID: 69e772de47f34ce924cf31e7. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Size":"7B"}. This record does not assert independent reproduction. ### Qwen2.5-7B-Instruct + SFT BoN@3 Accuracy: 15.36. Result ID: 69e772de47f34ce924cf31e6. [Reporting paper]() Recorded evaluation configuration: {"Model Size":"7B"}. This record does not assert independent reproduction. ### Qwen2.5-7B-Instruct BoN@3 Accuracy: 14.76. Result ID: 69e772de47f34ce924cf31e5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"7B"}. This record does not assert independent reproduction. ### Qwen2.5-7B-Instruct + ToRL BoN@3 Accuracy: 12.56. Result ID: 69e772de47f34ce924cf31e8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"7B"}. This record does not assert independent reproduction. ### Qwen2.5-3B-Instruct + GRPO BoN@3 Accuracy: 12.13. Result ID: 69e772de47f34ce924cf31e0. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Size":"3B"}. This record does not assert independent reproduction. ### Qwen2.5-3B-Instruct + ToRL BoN@3 Accuracy: 11.63. Result ID: 69e772de47f34ce924cf31e1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"3B"}. This record does not assert independent reproduction. ### GeometryZero-3B BoN@3 Accuracy: 11.3. Result ID: 69e772de47f34ce924cf31e2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"3B"}. This record does not assert independent reproduction. ### Qwen2.5-3B-Instruct BoN@3 Accuracy: 10.53. Result ID: 69e772de47f34ce924cf31de. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"3B"}. This record does not assert independent reproduction. ### Qwen2.5-3B-Instruct + SFT BoN@3 Accuracy: 10.2. Result ID: 69e772de47f34ce924cf31df. [Reporting paper]() Recorded evaluation configuration: {"Model Size":"3B"}. This record does not assert independent reproduction. ### GeometryZero-1.5B BoN@3 Accuracy: 6.95. Result ID: 69e772de47f34ce924cf31dd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"1.5B"}. This record does not assert independent reproduction. ### G-Llava-7B BoN@3 Accuracy: 6.23. Result ID: 69e772de47f34ce924cf31e3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"7B"}. This record does not assert independent reproduction. ### Qwen2.5-1.5B-Instruct + GRPO BoN@3 Accuracy: 5.76. Result ID: 69e772de47f34ce924cf31db. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Size":"1.5B"}. This record does not assert independent reproduction. ### Qwen2.5-1.5B-Instruct + ToRL BoN@3 Accuracy: 5.26. Result ID: 69e772de47f34ce924cf31dc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"1.5B"}. This record does not assert independent reproduction. ### GNS-Llava-1.5-7B BoN@3 Accuracy: 5.21. Result ID: 69e772de47f34ce924cf31e4. [Reporting paper]() Recorded evaluation configuration: {"Model Size":"7B"}. This record does not assert independent reproduction. ### Qwen2.5-1.5B-Instruct + SFT BoN@3 Accuracy: 4.8. Result ID: 69e772de47f34ce924cf31da. [Reporting paper]() Recorded evaluation configuration: {"Model Size":"1.5B"}. This record does not assert independent reproduction. ### Qwen2.5-1.5B-Instruct BoN@3 Accuracy: 4.2. Result ID: 69e772de47f34ce924cf31d9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"1.5B"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.