# AI-Text Detection on AI-Generated Text (Held-out Set) Source: [AI-Text Detection on AI-Generated Text (Held-out Set)]() Collection index: [llms.txt]() Data updated: 2026-06-01T06:44:10.028Z Evaluation of AI-Text Detection models specifically on the held-out set of the AI-Generated Text dataset, utilizing specialized metrics like AI-detection rate (Aligned, Pangram) and changes across various PASTA perturbation sets. 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-Text Detection | | Dataset (including any stated split) | AI-Generated Text (Held-out Set) | | Metric | AI-detection Rate (Aligned, Pangram) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 11 of 11 | | 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 - [AI-detection Rate (Aligned, Pangram)](): Higher is better; unit: PERCENT. - [Change in AI-detection Rate (PASTA, Pangram)](): Lower is better; unit: SCALAR. - [Change in AI-detection Rate (PASTA, Binoculars)](): Lower is better; unit: SCALAR. - [Change in AI-detection Rate (PASTA, ImBD)](): Lower is better; unit: SCALAR. - [Change in AI-detection Rate (PASTA, RAIDAR)](): Lower is better; unit: SCALAR. ## Recorded results | Method | AI-detection Rate (Aligned, Pangram) | Result date | Report paper | | --- | --- | --- | --- | | PASTA | 100 | Not recorded | 6a1d1207dbd7150506f83f30 | | PASTA | 99.8 | Not recorded | 6a1d1207dbd7150506f83f30 | | PASTA | 99.2 | Not recorded | 6a1d1207dbd7150506f83f30 | | PASTA | 99 | Not recorded | 6a1d1207dbd7150506f83f30 | | PASTA | 99 | Not recorded | 6a1d1207dbd7150506f83f30 | | PASTA | 99 | Not recorded | 6a1d1207dbd7150506f83f30 | | PASTA | 98.8 | Not recorded | 6a1d1207dbd7150506f83f30 | | PASTA | 98.6 | Not recorded | 6a1d1207dbd7150506f83f30 | | PASTA | 98.2 | Not recorded | 6a1d1207dbd7150506f83f30 | | PASTA | 97.4 | Not recorded | 6a1d1207dbd7150506f83f30 | | PASTA | 95.6 | Not recorded | 6a1d1207dbd7150506f83f30 | ### PASTA AI-detection Rate (Aligned, Pangram): 100. Result ID: 6a1d2989fbc373592c15d809. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Gemma-2-9B"}. This record does not assert independent reproduction. ### PASTA AI-detection Rate (Aligned, Pangram): 99.8. Result ID: 6a1d2989fbc373592c15d80f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B-Inst."}. This record does not assert independent reproduction. ### PASTA AI-detection Rate (Aligned, Pangram): 99.2. Result ID: 6a1d2989fbc373592c15d811. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-0.5B"}. This record does not assert independent reproduction. ### PASTA AI-detection Rate (Aligned, Pangram): 99. Result ID: 6a1d2989fbc373592c15d80c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-1.5B"}. This record does not assert independent reproduction. ### PASTA AI-detection Rate (Aligned, Pangram): 99. Result ID: 6a1d2989fbc373592c15d810. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Mistral-7B-v0.3"}. This record does not assert independent reproduction. ### PASTA AI-detection Rate (Aligned, Pangram): 99. Result ID: 6a1d2989fbc373592c15d812. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Tulu-3-8B-RLVR"}. This record does not assert independent reproduction. ### PASTA AI-detection Rate (Aligned, Pangram): 98.8. Result ID: 6a1d2989fbc373592c15d80b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-14B"}. This record does not assert independent reproduction. ### PASTA AI-detection Rate (Aligned, Pangram): 98.6. Result ID: 6a1d2989fbc373592c15d80d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### PASTA AI-detection Rate (Aligned, Pangram): 98.2. Result ID: 6a1d2989fbc373592c15d80a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-32B"}. This record does not assert independent reproduction. ### PASTA AI-detection Rate (Aligned, Pangram): 97.4. Result ID: 6a1d2989fbc373592c15d80e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-3B"}. This record does not assert independent reproduction. ### PASTA AI-detection Rate (Aligned, Pangram): 95.6. Result ID: 6a1d2989fbc373592c15d813. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"OLMo-3-7B-Inst."}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.