# Language Modeling on CNN news En (test)

Source: [Language Modeling on CNN news En (test)](<https://www.sota2.com/research/sota/language-modeling-on-cnn-news-en-test>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T05:37:24.441Z

Evaluation of Language Modeling performance on the test split of the CNN news English dataset, measured using Bits Per Character (BPC) variants.

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 | Language Modeling |
| Dataset (including any stated split) | CNN news En (test) |
| Metric | BPC (df) |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 2 of 2 |
| 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

- [BPC (df)](<https://www.sota2.com/research/sota/language-modeling-on-cnn-news-en-test.md?metric=BPC%20(df)>): Lower is better; unit: SCALAR.

- [BPC (is)](<https://www.sota2.com/research/sota/language-modeling-on-cnn-news-en-test.md?metric=BPC%20(is)>): Lower is better; unit: SCALAR.

- [BPC Gap](<https://www.sota2.com/research/sota/language-modeling-on-cnn-news-en-test.md?metric=BPC%20Gap>): Lower is better; unit: SCALAR.

- [% Relative BPC Gap](<https://www.sota2.com/research/sota/language-modeling-on-cnn-news-en-test.md?metric=%25%20Relative%20BPC%20Gap>): Lower is better; unit: PERCENT.

- [% ND](<https://www.sota2.com/research/sota/language-modeling-on-cnn-news-en-test.md?metric=%25%20ND>): Lower is better; unit: PERCENT.

## Recorded results

| Method | BPC (df) | Result date | Report paper |
| --- | --- | --- | --- |
| BLOOM | 0.8499 | Not recorded | 68fa6fee1e209131cea48381 |
| GPT-2 | 0.9421 | Not recorded | 68fa6fee1e209131cea48381 |

### BLOOM

BPC (df): 0.8499. Result ID: 699102ea3a6c3148490501f8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6fee1e209131cea48381.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6fdd1e209131cea47c06.md>)

Recorded evaluation configuration: {"Parameters":"1.7B"}. This record does not assert independent reproduction.

### GPT-2

BPC (df): 0.9421. Result ID: 699102ea3a6c3148490501f7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6fee1e209131cea48381.md>)

Recorded evaluation configuration: {"Parameters":"125M"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.