# Next Token Prediction on OpenWebText (PPL, Accuracy) Source: [Next Token Prediction on OpenWebText (PPL, Accuracy)]() Collection index: [llms.txt]() Data updated: 2026-04-01T07:02:21.126Z Evaluation of language model performance on the OpenWebText dataset using Next Token Prediction, measured by PPL 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 | Next Token Prediction | | Dataset (including any stated split) | OpenWebText | | Metric | PPL | | Unit | SCALAR | | Direction | Lower is better | | Results loaded | 30 of 30 | | 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 - [PPL](): Lower is better; unit: SCALAR. - [Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | PPL | Result date | Report paper | | --- | --- | --- | --- | | Concepts | 18.68 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Concepts | 18.68 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Concepts | 19.19 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Concepts | 20.26 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Baseline | 22.26 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Baseline | 22.42 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Concepts | 22.8 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Concepts | 23.22 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Concepts | 24.12 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Concepts | 25.62 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Baseline | 28.39 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Baseline | 28.43 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Concepts | 30.47 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Concepts | 31.28 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Concepts | 32.65 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Concepts | 34.84 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Baseline | 38.56 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Baseline | 38.71 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Noise | 39.49 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Noise | 47.92 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Noise | 62.75 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Noise | 78.4 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Noise | 99.13 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Noise | 130.08 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Noise | 203.15 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Noise | 264.21 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Noise | 374.05 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Noise | 28643.16 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Noise | 28898.47 | Not recorded | 69cca67e0636e5bf4afcbb28 | | Noise | 35501.71 | Not recorded | 69cca67e0636e5bf4afcbb28 | ### Concepts PPL: 18.68. Result ID: 69ccc2f547f34ce924b5f71b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama 3.1 8B","lambda":"0.75"}. This record does not assert independent reproduction. ### Concepts PPL: 18.68. Result ID: 69ccc2f547f34ce924b5f71c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama 3.1 8B","lambda":"1.0"}. This record does not assert independent reproduction. ### Concepts PPL: 19.19. Result ID: 69ccc2f547f34ce924b5f71a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama 3.1 8B","lambda":"0.5"}. This record does not assert independent reproduction. ### Concepts PPL: 20.26. Result ID: 69ccc2f547f34ce924b5f719. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama 3.1 8B","lambda":"0.25"}. This record does not assert independent reproduction. ### Baseline PPL: 22.26. Result ID: 69ccc2f547f34ce924b5f717. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama 3.1 8B","lambda":"Base"}. This record does not assert independent reproduction. ### Baseline PPL: 22.42. Result ID: 69ccc2f547f34ce924b5f718. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama 3.1 8B","lambda":"0.0"}. This record does not assert independent reproduction. ### Concepts PPL: 22.8. Result ID: 69ccc2f547f34ce924b5f712. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama 3.2 3B","lambda":"1.0"}. This record does not assert independent reproduction. ### Concepts PPL: 23.22. Result ID: 69ccc2f547f34ce924b5f711. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama 3.2 3B","lambda":"0.75"}. This record does not assert independent reproduction. ### Concepts PPL: 24.12. Result ID: 69ccc2f547f34ce924b5f710. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama 3.2 3B","lambda":"0.5"}. This record does not assert independent reproduction. ### Concepts PPL: 25.62. Result ID: 69ccc2f547f34ce924b5f70f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama 3.2 3B","lambda":"0.25"}. This record does not assert independent reproduction. ### Baseline PPL: 28.39. Result ID: 69ccc2f547f34ce924b5f70e. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama 3.2 3B","lambda":"0.0"}. This record does not assert independent reproduction. ### Baseline PPL: 28.43. Result ID: 69ccc2f547f34ce924b5f70d. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama 3.2 3B","lambda":"Base"}. This record does not assert independent reproduction. ### Concepts PPL: 30.47. Result ID: 69ccc2f547f34ce924b5f708. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama 3.2 1B","lambda":"1.0"}. This record does not assert independent reproduction. ### Concepts PPL: 31.28. Result ID: 69ccc2f547f34ce924b5f707. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama 3.2 1B","lambda":"0.75"}. This record does not assert independent reproduction. ### Concepts PPL: 32.65. Result ID: 69ccc2f547f34ce924b5f706. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama 3.2 1B","lambda":"0.5"}. This record does not assert independent reproduction. ### Concepts PPL: 34.84. Result ID: 69ccc2f547f34ce924b5f705. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama 3.2 1B","lambda":"0.25"}. This record does not assert independent reproduction. ### Baseline PPL: 38.56. Result ID: 69ccc2f547f34ce924b5f703. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama 3.2 1B","lambda":"Base"}. This record does not assert independent reproduction. ### Baseline PPL: 38.71. Result ID: 69ccc2f547f34ce924b5f704. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama 3.2 1B","lambda":"0.0"}. This record does not assert independent reproduction. ### Noise PPL: 39.49. Result ID: 69ccc2f547f34ce924b5f71d. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama 3.1 8B","lambda":"0.25"}. This record does not assert independent reproduction. ### Noise PPL: 47.92. Result ID: 69ccc2f547f34ce924b5f713. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama 3.2 3B","lambda":"0.25"}. This record does not assert independent reproduction. ### Noise PPL: 62.75. Result ID: 69ccc2f547f34ce924b5f709. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama 3.2 1B","lambda":"0.25"}. This record does not assert independent reproduction. ### Noise PPL: 78.4. Result ID: 69ccc2f547f34ce924b5f71e. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama 3.1 8B","lambda":"0.5"}. This record does not assert independent reproduction. ### Noise PPL: 99.13. Result ID: 69ccc2f547f34ce924b5f714. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama 3.2 3B","lambda":"0.5"}. This record does not assert independent reproduction. ### Noise PPL: 130.08. Result ID: 69ccc2f547f34ce924b5f70a. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama 3.2 1B","lambda":"0.5"}. This record does not assert independent reproduction. ### Noise PPL: 203.15. Result ID: 69ccc2f547f34ce924b5f71f. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama 3.1 8B","lambda":"0.75"}. This record does not assert independent reproduction. ### Noise PPL: 264.21. Result ID: 69ccc2f547f34ce924b5f715. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama 3.2 3B","lambda":"0.75"}. This record does not assert independent reproduction. ### Noise PPL: 374.05. Result ID: 69ccc2f547f34ce924b5f70b. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama 3.2 1B","lambda":"0.75"}. This record does not assert independent reproduction. ### Noise PPL: 28643.16. Result ID: 69ccc2f547f34ce924b5f70c. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama 3.2 1B","lambda":"1.0"}. This record does not assert independent reproduction. ### Noise PPL: 28898.47. Result ID: 69ccc2f547f34ce924b5f716. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama 3.2 3B","lambda":"1.0"}. This record does not assert independent reproduction. ### Noise PPL: 35501.71. Result ID: 69ccc2f547f34ce924b5f720. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama 3.1 8B","lambda":"1.0"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.