# AI-generated text detection on M4GT-Bench Wikipedia (test) Source: [AI-generated text detection on M4GT-Bench Wikipedia (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:58:43.484Z Evaluation of AI-generated text detection models on the Wikipedia test split of the M4GT-Bench dataset, reporting Precision, Recall, F1, and 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 | AI-generated text detection | | Dataset (including any stated split) | M4GT-Bench Wikipedia (test) | | Metric | Precision | | 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 - [Precision](): Higher is better; unit: PERCENT. - [Recall](): Higher is better; unit: PERCENT. - [F1 Score](): Higher is better; unit: PERCENT. - [Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Precision | Result date | Report paper | | --- | --- | --- | --- | | XLM-R | 60.04 | Not recorded | 6995478c09c33900fcd6d5da | | Stylistic-SVM | 47.94 | Not recorded | 6995478c09c33900fcd6d5da | | NELA-SVM | 46.38 | Not recorded | 6995478c09c33900fcd6d5da | | RoBERTa | 46.37 | Not recorded | 6995478c09c33900fcd6d5da | | DependencyAI | 44.45 | Not recorded | 6995478c09c33900fcd6d5da | | GLTR-LR | 41.62 | Not recorded | 6995478c09c33900fcd6d5da | | GLTR-SVM | 34.86 | Not recorded | 6995478c09c33900fcd6d5da | ### XLM-R Precision: 60.04. Result ID: 699ad87f5698665ec467eb9a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Protocol":"Leave-one-domain-out"}. This record does not assert independent reproduction. ### Stylistic-SVM Precision: 47.94. Result ID: 699ad87f5698665ec467eb9d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Protocol":"Leave-one-domain-out"}. This record does not assert independent reproduction. ### NELA-SVM Precision: 46.38. Result ID: 699ad87f5698665ec467eb9e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Protocol":"Leave-one-domain-out"}. This record does not assert independent reproduction. ### RoBERTa Precision: 46.37. Result ID: 699ad87f5698665ec467eb99. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Protocol":"Leave-one-domain-out"}. This record does not assert independent reproduction. ### DependencyAI Precision: 44.45. Result ID: 699ad87f5698665ec467eb9f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Protocol":"Leave-one-domain-out"}. This record does not assert independent reproduction. ### GLTR-LR Precision: 41.62. Result ID: 699ad87f5698665ec467eb9b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Protocol":"Leave-one-domain-out"}. This record does not assert independent reproduction. ### GLTR-SVM Precision: 34.86. Result ID: 699ad87f5698665ec467eb9c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Protocol":"Leave-one-domain-out"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.