# Edit Inspectors Questions on UltraEdit Source: [Edit Inspectors Questions on UltraEdit]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:27:00.127Z Evaluation of model performance on the UltraEdit dataset, assessing the quality of edits based on inspector questions using metrics like Accuracy, Contextual Consistency, and Technical Precision. 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 | Edit Inspectors Questions | | Dataset (including any stated split) | UltraEdit | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 7 of 7 | | 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. - [Contextual Consistency](): Higher is better; unit: PERCENT. - [Technical Precision](): Higher is better; unit: PERCENT. - [Artifacts Score](): Higher is better; unit: PERCENT. - [Caption Acc (Diff)](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | GPT-4 | 63 | Not recorded | 68fa79651e209131cea7d6a6 | | Qwen2.5 VL | 62.1 | Not recorded | 68fa79651e209131cea7d6a6 | | GPT-4 Turbo | 57.5 | Not recorded | 68fa79651e209131cea7d6a6 | | InternVL3 | 54.4 | Not recorded | 68fa79651e209131cea7d6a6 | | LLaVA | 54.3 | Not recorded | 68fa79651e209131cea7d6a6 | | LLaVA | 52.6 | Not recorded | 68fa79651e209131cea7d6a6 | | GPT-4o | 51.8 | Not recorded | 68fa79651e209131cea7d6a6 | ### GPT-4 Accuracy: 63. Result ID: 698b4b8e8d67a41d0a6ce153. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen2.5 VL Accuracy: 62.1. Result ID: 698b4b8e8d67a41d0a6ce156. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### GPT-4 Turbo Accuracy: 57.5. Result ID: 698b4b8e8d67a41d0a6ce155. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### InternVL3 Accuracy: 54.4. Result ID: 698b4b8e8d67a41d0a6ce157. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### LLaVA Accuracy: 54.3. Result ID: 698b4b8e8d67a41d0a6ce159. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Training Protocol":"Supervised"}. This record does not assert independent reproduction. ### LLaVA Accuracy: 52.6. Result ID: 698b4b8e8d67a41d0a6ce158. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### GPT-4o Accuracy: 51.8. Result ID: 698b4b8e8d67a41d0a6ce154. [Reporting paper]() [Method origin paper]() [Implementation]() 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.