# Fine-grained Understanding on UHR-Micro (test) Source: [Fine-grained Understanding on UHR-Micro (test)]() Collection index: [llms.txt]() Data updated: 2026-05-13T10:48:27.152Z Evaluation of model performance on the test split of the UHR-Micro dataset focusing on Fine-grained Understanding tasks. 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 | Fine-grained Understanding | | Dataset (including any stated split) | UHR-Micro (test) | | Metric | OC Score | | Unit | SCALAR | | Direction | Higher is better | | Results loaded | 16 of 16 | | 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 - [OC Score](): Higher is better; unit: SCALAR. - [FGR Score](): Higher is better; unit: SCALAR. - [RS Score](): Higher is better; unit: SCALAR. - [CS Score](): Higher is better; unit: SCALAR. - [Average Score](): Higher is better; unit: SCALAR. ## Recorded results | Method | OC Score | Result date | Report paper | | --- | --- | --- | --- | | MAP-Agent | 64.32 | Not recorded | 6a0405dedbc836b87d8278bd | | MAP-Agent | 62.09 | Not recorded | 6a0405dedbc836b87d8278bd | | Qwen2.5-VL-7B | 58.83 | Not recorded | 6a0405dedbc836b87d8278bd | | DeepEyes | 57.12 | Not recorded | 6a0405dedbc836b87d8278bd | | Gemini-2.5-Pro | 55.57 | Not recorded | 6a0405dedbc836b87d8278bd | | Qwen3-VL-235B | 55.06 | Not recorded | 6a0405dedbc836b87d8278bd | | InternVL3.5-38B | 54.37 | Not recorded | 6a0405dedbc836b87d8278bd | | Qwen3-VL-8B | 51.97 | Not recorded | 6a0405dedbc836b87d8278bd | | Qwen3-VL-30B | 51.97 | Not recorded | 6a0405dedbc836b87d8278bd | | Qwen2-VL-7B | 50.94 | Not recorded | 6a0405dedbc836b87d8278bd | | GPT-5.2 | 50.09 | Not recorded | 6a0405dedbc836b87d8278bd | | InternVL3.5-8B | 49.06 | Not recorded | 6a0405dedbc836b87d8278bd | | GeoChat | 47.17 | Not recorded | 6a0405dedbc836b87d8278bd | | GeoLLaVa-8k | 30.53 | Not recorded | 6a0405dedbc836b87d8278bd | | Claude 3.7 Sonnet | 26.59 | Not recorded | 6a0405dedbc836b87d8278bd | | GeoEyes | 1.72 | Not recorded | 6a0405dedbc836b87d8278bd | ### MAP-Agent OC Score: 64.32. Result ID: 6a0456e047f34ce924eafaac. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"base\_model":"Qwen3-VL-8B"}. This record does not assert independent reproduction. ### MAP-Agent OC Score: 62.09. Result ID: 6a0456e047f34ce924eafaab. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"base\_model":"InternVL3.5-8B"}. This record does not assert independent reproduction. ### Qwen2.5-VL-7B OC Score: 58.83. Result ID: 6a0456e047f34ce924eafa9e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### DeepEyes OC Score: 57.12. Result ID: 6a0456e047f34ce924eafaa7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Gemini-2.5-Pro OC Score: 55.57. Result ID: 6a0456e047f34ce924eafaa5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen3-VL-235B OC Score: 55.06. Result ID: 6a0456e047f34ce924eafaa1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### InternVL3.5-38B OC Score: 54.37. Result ID: 6a0456e047f34ce924eafaa3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen3-VL-8B OC Score: 51.97. Result ID: 6a0456e047f34ce924eafa9f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen3-VL-30B OC Score: 51.97. Result ID: 6a0456e047f34ce924eafaa0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen2-VL-7B OC Score: 50.94. Result ID: 6a0456e047f34ce924eafa9d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### GPT-5.2 OC Score: 50.09. Result ID: 6a0456e047f34ce924eafaa4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### InternVL3.5-8B OC Score: 49.06. Result ID: 6a0456e047f34ce924eafaa2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### GeoChat OC Score: 47.17. Result ID: 6a0456e047f34ce924eafaa8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### GeoLLaVa-8k OC Score: 30.53. Result ID: 6a0456e047f34ce924eafaa9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Claude 3.7 Sonnet OC Score: 26.59. Result ID: 6a0456e047f34ce924eafaa6. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### GeoEyes OC Score: 1.72. Result ID: 6a0456e047f34ce924eafaaa. [Reporting paper]() [Method origin paper]() 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.