# General Reasoning on MMStar (Score) Source: [General Reasoning on MMStar (Score)]() Collection index: [llms.txt]() Data updated: 2026-02-26T06:03:49.403Z Evaluation of model performance on the MMStar dataset for General Reasoning tasks, measured by Score. 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 | General Reasoning | | Dataset (including any stated split) | MMStar | | Metric | Score | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 32 of 32 | | 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 - [Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | Score | Result date | Report paper | | --- | --- | --- | --- | | ViLoMem | 69.2 | Not recorded | 698180f9a8b01e76cc8e2199 | | Vanilla RAG | 68.4 | Not recorded | 698180f9a8b01e76cc8e2199 | | PolarMem | 68.4 | Not recorded | 698180f9a8b01e76cc8e2199 | | ViLoMem | 68.1 | Not recorded | 698180f9a8b01e76cc8e2199 | | InternVL-3.5-8B | 68.1 | Not recorded | 698180f9a8b01e76cc8e2199 | | QWEN2.5-VL-32B | 67.8 | Not recorded | 698180f9a8b01e76cc8e2199 | | PolarMem | 67.2 | Not recorded | 698180f9a8b01e76cc8e2199 | | Vanilla RAG | 67.2 | Not recorded | 698180f9a8b01e76cc8e2199 | | PolarMem | 66.2 | Not recorded | 698180f9a8b01e76cc8e2199 | | ViLoMem | 65.7 | Not recorded | 698180f9a8b01e76cc8e2199 | | PolarMem | 65.1 | Not recorded | 698180f9a8b01e76cc8e2199 | | InternVL-3.5-4B | 65.1 | Not recorded | 698180f9a8b01e76cc8e2199 | | Vanilla RAG | 64.9 | Not recorded | 698180f9a8b01e76cc8e2199 | | ViLoMem | 64.4 | Not recorded | 698180f9a8b01e76cc8e2199 | | QWEN2.5-VL-7B | 63.6 | Not recorded | 698180f9a8b01e76cc8e2199 | | ViLoMem | 63.6 | Not recorded | 698180f9a8b01e76cc8e2199 | | Vanilla RAG | 63.4 | Not recorded | 698180f9a8b01e76cc8e2199 | | PolarMem | 62.8 | Not recorded | 698180f9a8b01e76cc8e2199 | | PolarMem | 62.4 | Not recorded | 698180f9a8b01e76cc8e2199 | | PolarMem | 61.8 | Not recorded | 698180f9a8b01e76cc8e2199 | | ViLoMem | 61.6 | Not recorded | 698180f9a8b01e76cc8e2199 | | LLaVA-Next-Llama-3-8B | 61.5 | Not recorded | 698180f9a8b01e76cc8e2199 | | ViLoMem | 61.2 | Not recorded | 698180f9a8b01e76cc8e2199 | | Vanilla RAG | 60.8 | Not recorded | 698180f9a8b01e76cc8e2199 | | DeepSeek-VL2 | 60.7 | Not recorded | 698180f9a8b01e76cc8e2199 | | LLaVA-Next-Mistral-7B | 60.3 | Not recorded | 698180f9a8b01e76cc8e2199 | | Vanilla RAG | 59.7 | Not recorded | 698180f9a8b01e76cc8e2199 | | ViLoMem | 59 | Not recorded | 698180f9a8b01e76cc8e2199 | | Vanilla RAG | 58.8 | Not recorded | 698180f9a8b01e76cc8e2199 | | PolarMem | 57.8 | Not recorded | 698180f9a8b01e76cc8e2199 | | DeepSeek-VL2-Small | 56.9 | Not recorded | 698180f9a8b01e76cc8e2199 | | Vanilla RAG | 56.2 | Not recorded | 698180f9a8b01e76cc8e2199 | ### ViLoMem Score: 69.2. Result ID: 699613705698665ec452d1ac. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"InternVL-3.5-8B","Augmentation":"ViLoMem"}. This record does not assert independent reproduction. ### Vanilla RAG Score: 68.4. Result ID: 699613705698665ec452d1a3. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"QWEN2.5-VL-32B","Augmentation":"Vanilla RAG"}. This record does not assert independent reproduction. ### PolarMem Score: 68.4. Result ID: 699613705698665ec452d1ad. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"InternVL-3.5-8B","Augmentation":"PolarMem"}. This record does not assert independent reproduction. ### ViLoMem Score: 68.1. Result ID: 699613705698665ec452d1a4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"QWEN2.5-VL-32B","Augmentation":"ViLoMem"}. This record does not assert independent reproduction. ### InternVL-3.5-8B Score: 68.1. Result ID: 699613705698665ec452d1aa. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"InternVL-3.5-8B","Augmentation":"None"}. This record does not assert independent reproduction. ### QWEN2.5-VL-32B Score: 67.8. Result ID: 699613705698665ec452d1a2. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"QWEN2.5-VL-32B","Augmentation":"None"}. This record does not assert independent reproduction. ### PolarMem Score: 67.2. Result ID: 699613705698665ec452d1a5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"QWEN2.5-VL-32B","Augmentation":"PolarMem"}. This record does not assert independent reproduction. ### Vanilla RAG Score: 67.2. Result ID: 699613705698665ec452d1ab. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"InternVL-3.5-8B","Augmentation":"Vanilla RAG"}. This record does not assert independent reproduction. ### PolarMem Score: 66.2. Result ID: 699613705698665ec452d1a9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"InternVL-3.5-4B","Augmentation":"PolarMem"}. This record does not assert independent reproduction. ### ViLoMem Score: 65.7. Result ID: 699613705698665ec452d1a8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"InternVL-3.5-4B","Augmentation":"ViLoMem"}. This record does not assert independent reproduction. ### PolarMem Score: 65.1. Result ID: 699613705698665ec452d1a1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"QWEN2.5-VL-7B","Augmentation":"PolarMem"}. This record does not assert independent reproduction. ### InternVL-3.5-4B Score: 65.1. Result ID: 699613705698665ec452d1a6. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"InternVL-3.5-4B","Augmentation":"None"}. This record does not assert independent reproduction. ### Vanilla RAG Score: 64.9. Result ID: 699613705698665ec452d1a7. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"InternVL-3.5-4B","Augmentation":"Vanilla RAG"}. This record does not assert independent reproduction. ### ViLoMem Score: 64.4. Result ID: 699613705698665ec452d1a0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"QWEN2.5-VL-7B","Augmentation":"ViLoMem"}. This record does not assert independent reproduction. ### QWEN2.5-VL-7B Score: 63.6. Result ID: 699613705698665ec452d19e. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"QWEN2.5-VL-7B","Augmentation":"None"}. This record does not assert independent reproduction. ### ViLoMem Score: 63.6. Result ID: 699613705698665ec452d1bc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"LLaVA-Next-Llama-3-8B","Augmentation":"ViLoMem"}. This record does not assert independent reproduction. ### Vanilla RAG Score: 63.4. Result ID: 699613705698665ec452d19f. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"QWEN2.5-VL-7B","Augmentation":"Vanilla RAG"}. This record does not assert independent reproduction. ### PolarMem Score: 62.8. Result ID: 699613705698665ec452d1bd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"LLaVA-Next-Llama-3-8B","Augmentation":"PolarMem"}. This record does not assert independent reproduction. ### PolarMem Score: 62.4. Result ID: 699613705698665ec452d1b5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"DeepSeek-VL2","Augmentation":"PolarMem"}. This record does not assert independent reproduction. ### PolarMem Score: 61.8. Result ID: 699613705698665ec452d1b9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"LLaVA-Next-Mistral-7B","Augmentation":"PolarMem"}. This record does not assert independent reproduction. ### ViLoMem Score: 61.6. Result ID: 699613705698665ec452d1b4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"DeepSeek-VL2","Augmentation":"ViLoMem"}. This record does not assert independent reproduction. ### LLaVA-Next-Llama-3-8B Score: 61.5. Result ID: 699613705698665ec452d1ba. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"LLaVA-Next-Llama-3-8B","Augmentation":"None"}. This record does not assert independent reproduction. ### ViLoMem Score: 61.2. Result ID: 699613705698665ec452d1b8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"LLaVA-Next-Mistral-7B","Augmentation":"ViLoMem"}. This record does not assert independent reproduction. ### Vanilla RAG Score: 60.8. Result ID: 699613705698665ec452d1bb. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"LLaVA-Next-Llama-3-8B","Augmentation":"Vanilla RAG"}. This record does not assert independent reproduction. ### DeepSeek-VL2 Score: 60.7. Result ID: 699613705698665ec452d1b2. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"DeepSeek-VL2","Augmentation":"None"}. This record does not assert independent reproduction. ### LLaVA-Next-Mistral-7B Score: 60.3. Result ID: 699613705698665ec452d1b6. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"LLaVA-Next-Mistral-7B","Augmentation":"None"}. This record does not assert independent reproduction. ### Vanilla RAG Score: 59.7. Result ID: 699613705698665ec452d1b3. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"DeepSeek-VL2","Augmentation":"Vanilla RAG"}. This record does not assert independent reproduction. ### ViLoMem Score: 59. Result ID: 699613705698665ec452d1b0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"DeepSeek-VL2-Small","Augmentation":"ViLoMem"}. This record does not assert independent reproduction. ### Vanilla RAG Score: 58.8. Result ID: 699613705698665ec452d1b7. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"LLaVA-Next-Mistral-7B","Augmentation":"Vanilla RAG"}. This record does not assert independent reproduction. ### PolarMem Score: 57.8. Result ID: 699613705698665ec452d1b1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"DeepSeek-VL2-Small","Augmentation":"PolarMem"}. This record does not assert independent reproduction. ### DeepSeek-VL2-Small Score: 56.9. Result ID: 699613705698665ec452d1ae. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"DeepSeek-VL2-Small","Augmentation":"None"}. This record does not assert independent reproduction. ### Vanilla RAG Score: 56.2. Result ID: 699613705698665ec452d1af. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"DeepSeek-VL2-Small","Augmentation":"Vanilla RAG"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.