# Visual Mathematical Reasoning on WeMath strict Source: [Visual Mathematical Reasoning on WeMath strict]() Collection index: [llms.txt]() Data updated: 2026-05-15T07:09:43.832Z Evaluation of model performance on visual mathematical reasoning tasks using the strict split/version of the WeMath dataset, typically measured by Accuracy and Average 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 | Visual Mathematical Reasoning | | Dataset (including any stated split) | WeMath strict | | 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. - [Average Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | AdaVaR-7B | 44.8 | Not recorded | 68fa7ba61e209131cea899f4 | | OVR-7B | 44.6 | Not recorded | 68fa7ba61e209131cea899f4 | | GPT-4o | 42.9 | Not recorded | 68fa7ba61e209131cea899f4 | | VLAA-Thinker-7B | 41.5 | Not recorded | 68fa7ba61e209131cea899f4 | | InternVL3-8B | 37.5 | Not recorded | 68fa7ba61e209131cea899f4 | | MM-Eureka | 36.9 | Not recorded | 68fa7ba61e209131cea899f4 | | ViGoRL-7B | 36.3 | Not recorded | 68fa7ba61e209131cea899f4 | | VLAA-Thinker-3B | 33.8 | Not recorded | 68fa7ba61e209131cea899f4 | | AdaVaR-3B | 33.8 | Not recorded | 68fa7ba61e209131cea899f4 | | Chain-of-Focus | 33.1 | Not recorded | 68fa7ba61e209131cea899f4 | | DeepEyes | 32.7 | Not recorded | 68fa7ba61e209131cea899f4 | | Orsta-7B | 31.8 | Not recorded | 68fa7ba61e209131cea899f4 | | Qwen2.5-VL-7B | 31.2 | Not recorded | 68fa7ba61e209131cea899f4 | | LMM-R1 | 29.2 | Not recorded | 68fa7ba61e209131cea899f4 | | GRIT-3B | 29.1 | Not recorded | 68fa7ba61e209131cea899f4 | | ViGoRL-3B | 28.5 | Not recorded | 68fa7ba61e209131cea899f4 | | Qwen2.5-VL-3B | 21.2 | Not recorded | 68fa7ba61e209131cea899f4 | | LLaVA-OV-7B | 17.7 | Not recorded | 68fa7ba61e209131cea899f4 | ### AdaVaR-7B Accuracy: 44.8. Result ID: 6a06c6b347f34ce924ede227. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Reasoning Models","Base Model":"Qwen2.5-VL-7B","Parameter Size":"7B","Reasoning Mode":"Adaptive"}. This record does not assert independent reproduction. ### OVR-7B Accuracy: 44.6. Result ID: 6a06c6b347f34ce924ede222. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Reasoning Models","Base Model":"Qwen2.5-VL-7B","Parameter Size":"7B","Reasoning Mode":"Text-based"}. This record does not assert independent reproduction. ### GPT-4o Accuracy: 42.9. Result ID: 6a06c6b347f34ce924ede216. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"SFT & Close-source Models"}. This record does not assert independent reproduction. ### VLAA-Thinker-7B Accuracy: 41.5. Result ID: 6a06c6b347f34ce924ede220. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Reasoning Models","Base Model":"Qwen2.5-VL-7B","Parameter Size":"7B","Reasoning Mode":"Text-based"}. This record does not assert independent reproduction. ### InternVL3-8B Accuracy: 37.5. Result ID: 6a06c6b347f34ce924ede218. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"SFT & Close-source Models","Parameter Size":"8B"}. This record does not assert independent reproduction. ### MM-Eureka Accuracy: 36.9. Result ID: 6a06c6b347f34ce924ede221. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Reasoning Models","Base Model":"Qwen2.5-VL-7B","Parameter Size":"7B"}. This record does not assert independent reproduction. ### ViGoRL-7B Accuracy: 36.3. Result ID: 6a06c6b347f34ce924ede226. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Reasoning Models","Base Model":"Qwen2.5-VL-7B","Parameter Size":"7B"}. This record does not assert independent reproduction. ### VLAA-Thinker-3B Accuracy: 33.8. Result ID: 6a06c6b347f34ce924ede21b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Reasoning Models","Base Model":"Qwen2.5-VL-3B","Parameter Size":"3B","Reasoning Mode":"Text-based"}. This record does not assert independent reproduction. ### AdaVaR-3B Accuracy: 33.8. Result ID: 6a06c6b347f34ce924ede21e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Reasoning Models","Base Model":"Qwen2.5-VL-3B","Parameter Size":"3B","Reasoning Mode":"Adaptive"}. This record does not assert independent reproduction. ### Chain-of-Focus Accuracy: 33.1. Result ID: 6a06c6b347f34ce924ede225. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Reasoning Models","Base Model":"Qwen2.5-VL-7B","Parameter Size":"7B"}. This record does not assert independent reproduction. ### DeepEyes Accuracy: 32.7. Result ID: 6a06c6b347f34ce924ede224. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Reasoning Models","Base Model":"Qwen2.5-VL-7B","Parameter Size":"7B","Reasoning Mode":"Visually-grounded"}. This record does not assert independent reproduction. ### Orsta-7B Accuracy: 31.8. Result ID: 6a06c6b347f34ce924ede223. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Reasoning Models","Base Model":"Qwen2.5-VL-7B","Parameter Size":"7B"}. This record does not assert independent reproduction. ### Qwen2.5-VL-7B Accuracy: 31.2. Result ID: 6a06c6b347f34ce924ede21f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Reasoning Models","Base Model":"Qwen2.5-VL-7B","Parameter Size":"7B"}. This record does not assert independent reproduction. ### LMM-R1 Accuracy: 29.2. Result ID: 6a06c6b347f34ce924ede21a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Reasoning Models","Base Model":"Qwen2.5-VL-3B","Parameter Size":"3B"}. This record does not assert independent reproduction. ### GRIT-3B Accuracy: 29.1. Result ID: 6a06c6b347f34ce924ede21c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Reasoning Models","Base Model":"Qwen2.5-VL-3B","Parameter Size":"3B"}. This record does not assert independent reproduction. ### ViGoRL-3B Accuracy: 28.5. Result ID: 6a06c6b347f34ce924ede21d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Reasoning Models","Base Model":"Qwen2.5-VL-3B","Parameter Size":"3B"}. This record does not assert independent reproduction. ### Qwen2.5-VL-3B Accuracy: 21.2. Result ID: 6a06c6b347f34ce924ede219. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Reasoning Models","Base Model":"Qwen2.5-VL-3B","Parameter Size":"3B"}. This record does not assert independent reproduction. ### LLaVA-OV-7B Accuracy: 17.7. Result ID: 6a06c6b347f34ce924ede217. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Category":"SFT & Close-source Models","Parameter Size":"7B"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.