# Watermark Detection on ELI5 and ICLR papers IPI setting Source: [Watermark Detection on ELI5 and ICLR papers IPI setting]() Collection index: [llms.txt]() Data updated: 2026-04-07T07:15:45.782Z Evaluation of Watermark Detection performance on a combined dataset consisting of ELI5 examples and ICLR papers evaluated under the IPI setting, using metrics like ROC-AUC and True Positive Rate at specific False Positive thresholds (T@1%F, T@10%F). 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 | Watermark Detection | | Dataset (including any stated split) | ELI5 and ICLR papers IPI setting | | Metric | ROC-AUC | | Unit | SCALAR | | Direction | Higher is better | | Results loaded | 8 of 8 | | 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 - [ROC-AUC](): Higher is better; unit: SCALAR. - [TPR @ 1% FPR](): Higher is better; unit: PERCENT. - [TPR @ 10% FPR](): Higher is better; unit: PERCENT. ## Recorded results | Method | ROC-AUC | Result date | Report paper | | --- | --- | --- | --- | | Unicode ICW | 1 | Not recorded | 68fa78b31e209131cea79d68 | | Initials ICW | 0.997 | Not recorded | 68fa78b31e209131cea79d68 | | Lexical ICW | 0.997 | Not recorded | 68fa78b31e209131cea79d68 | | Acrostics ICW | 0.997 | Not recorded | 68fa78b31e209131cea79d68 | | Lexical ICW | 0.889 | Not recorded | 68fa78b31e209131cea79d68 | | Unicode ICW | 0.857 | Not recorded | 68fa78b31e209131cea79d68 | | Initials ICW | 0.62 | Not recorded | 68fa78b31e209131cea79d68 | | Acrostics ICW | 0.592 | Not recorded | 68fa78b31e209131cea79d68 | ### Unicode ICW ROC-AUC: 1. Result ID: 69d4aead47f34ce924bc06ad. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Language Model":"GPT-o3-mini"}. This record does not assert independent reproduction. ### Initials ICW ROC-AUC: 0.997. Result ID: 69d4aead47f34ce924bc06ae. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Language Model":"GPT-o3-mini"}. This record does not assert independent reproduction. ### Lexical ICW ROC-AUC: 0.997. Result ID: 69d4aead47f34ce924bc06af. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Language Model":"GPT-o3-mini"}. This record does not assert independent reproduction. ### Acrostics ICW ROC-AUC: 0.997. Result ID: 69d4aead47f34ce924bc06b0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Language Model":"GPT-o3-mini"}. This record does not assert independent reproduction. ### Lexical ICW ROC-AUC: 0.889. Result ID: 69d4aead47f34ce924bc06ab. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Language Model":"GPT-4o-mini"}. This record does not assert independent reproduction. ### Unicode ICW ROC-AUC: 0.857. Result ID: 69d4aead47f34ce924bc06a9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Language Model":"GPT-4o-mini"}. This record does not assert independent reproduction. ### Initials ICW ROC-AUC: 0.62. Result ID: 69d4aead47f34ce924bc06aa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Language Model":"GPT-4o-mini"}. This record does not assert independent reproduction. ### Acrostics ICW ROC-AUC: 0.592. Result ID: 69d4aead47f34ce924bc06ac. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Language Model":"GPT-4o-mini"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.