# Real-world Understanding on R-Bench Source: [Real-world Understanding on R-Bench]() Collection index: [llms.txt]() Data updated: 2026-02-27T06:02:30.922Z Evaluation of model performance on the Real-world Understanding task using the R-Bench dataset. 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 | Real-world Understanding | | Dataset (including any stated split) | R-Bench | | Metric | Distance Error | | Unit | SCALAR | | Direction | Lower is better | | Results loaded | 19 of 19 | | 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 - [Distance Error](): Lower is better; unit: SCALAR. ## Recorded results | Method | Distance Error | Result date | Report paper | | --- | --- | --- | --- | | Phi-3.5-Vision-4B | 55.5 | Not recorded | 68fa7bab1e209131cea89c7f | | InternVL2-1B | 55.6 | Not recorded | 68fa7bab1e209131cea89c7f | | InternVL2-2B | 56.8 | Not recorded | 68fa7bab1e209131cea89c7f | | InternVL2.5-1B | 59 | Not recorded | 68fa7bab1e209131cea89c7f | | InternVL2.5-2B | 62.2 | Not recorded | 68fa7bab1e209131cea89c7f | | Qwen2-VL-7B | 64 | Not recorded | 68fa7bab1e209131cea89c7f | | InternVL2-4B | 64.5 | Not recorded | 68fa7bab1e209131cea89c7f | | GPT-4V | 65.6 | Not recorded | 68fa7bab1e209131cea89c7f | | InternVL2.5-4B | 66.1 | Not recorded | 68fa7bab1e209131cea89c7f | | InternVL2-8B | 67.9 | Not recorded | 68fa7bab1e209131cea89c7f | | InternVL-Chat-V1.5 | 67.9 | Not recorded | 68fa7bab1e209131cea89c7f | | InternVL2.5-8B | 70.1 | Not recorded | 68fa7bab1e209131cea89c7f | | InternVL2-26B | 70.1 | Not recorded | 68fa7bab1e209131cea89c7f | | InternVL2.5-38B | 72.1 | Not recorded | 68fa7bab1e209131cea89c7f | | InternVL2.5-26B | 72.9 | Not recorded | 68fa7bab1e209131cea89c7f | | InternVL2-40B | 73.3 | Not recorded | 68fa7bab1e209131cea89c7f | | InternVL2-Llama3-76B | 74.1 | Not recorded | 68fa7bab1e209131cea89c7f | | InternVL2.5-78B | 77.2 | Not recorded | 68fa7bab1e209131cea89c7f | | GPT-4o-20240513 | 77.7 | Not recorded | 68fa7bab1e209131cea89c7f | ### Phi-3.5-Vision-4B Distance Error: 55.5. Result ID: 69a1335f35dfcabb00e5726d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Name":"Phi-3.5-Vision-4B"}. This record does not assert independent reproduction. ### InternVL2-1B Distance Error: 55.6. Result ID: 69a1335f35dfcabb00e57269. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Name":"InternVL2-1B"}. This record does not assert independent reproduction. ### InternVL2-2B Distance Error: 56.8. Result ID: 69a1335f35dfcabb00e5726b. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Name":"InternVL2-2B"}. This record does not assert independent reproduction. ### InternVL2.5-1B Distance Error: 59. Result ID: 69a1335f35dfcabb00e5726a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Name":"InternVL2.5-1B"}. This record does not assert independent reproduction. ### InternVL2.5-2B Distance Error: 62.2. Result ID: 69a1335f35dfcabb00e5726c. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Name":"InternVL2.5-2B"}. This record does not assert independent reproduction. ### Qwen2-VL-7B Distance Error: 64. Result ID: 69a1335f35dfcabb00e57270. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Name":"Qwen2-VL-7B"}. This record does not assert independent reproduction. ### InternVL2-4B Distance Error: 64.5. Result ID: 69a1335f35dfcabb00e5726e. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Name":"InternVL2-4B"}. This record does not assert independent reproduction. ### GPT-4V Distance Error: 65.6. Result ID: 69a1335f35dfcabb00e57278. [Reporting paper]() Recorded evaluation configuration: {"Model Name":"GPT-4V"}. This record does not assert independent reproduction. ### InternVL2.5-4B Distance Error: 66.1. Result ID: 69a1335f35dfcabb00e5726f. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Name":"InternVL2.5-4B"}. This record does not assert independent reproduction. ### InternVL2-8B Distance Error: 67.9. Result ID: 69a1335f35dfcabb00e57271. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Name":"InternVL2-8B"}. This record does not assert independent reproduction. ### InternVL-Chat-V1.5 Distance Error: 67.9. Result ID: 69a1335f35dfcabb00e57273. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Name":"InternVL-Chat-V1.5"}. This record does not assert independent reproduction. ### InternVL2.5-8B Distance Error: 70.1. Result ID: 69a1335f35dfcabb00e57272. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Name":"InternVL2.5-8B"}. This record does not assert independent reproduction. ### InternVL2-26B Distance Error: 70.1. Result ID: 69a1335f35dfcabb00e57274. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Name":"InternVL2-26B"}. This record does not assert independent reproduction. ### InternVL2.5-38B Distance Error: 72.1. Result ID: 69a1335f35dfcabb00e57277. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Name":"InternVL2.5-38B"}. This record does not assert independent reproduction. ### InternVL2.5-26B Distance Error: 72.9. Result ID: 69a1335f35dfcabb00e57275. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Name":"InternVL2.5-26B"}. This record does not assert independent reproduction. ### InternVL2-40B Distance Error: 73.3. Result ID: 69a1335f35dfcabb00e57276. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Name":"InternVL2-40B"}. This record does not assert independent reproduction. ### InternVL2-Llama3-76B Distance Error: 74.1. Result ID: 69a1335f35dfcabb00e5727a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Name":"InternVL2-Llama3-76B"}. This record does not assert independent reproduction. ### InternVL2.5-78B Distance Error: 77.2. Result ID: 69a1335f35dfcabb00e5727b. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Name":"InternVL2.5-78B"}. This record does not assert independent reproduction. ### GPT-4o-20240513 Distance Error: 77.7. Result ID: 69a1335f35dfcabb00e57279. [Reporting paper]() Recorded evaluation configuration: {"Model Name":"GPT-4o-20240513"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.