Document-level Chinese machine-generated-text detection data, plus English renderings of paired Chinese Wikipedia articles, plus 17 demonstration cases. Counts below are measured on this release. There is no train/test split in the files. The full split is the whole release. A paper may describe a document-level 9:1 protocol; the seed is not stored here, so this repository does not invent one. Filter with the subset column. News CJK totals: human 103,787,963; AI 243,603,638. Wiki-Claude CJK totals: human 8,115,726; AI 2,877,254. Literary works, translations, news articles, and the paper excerpt in examples remain under their original copyright. See LICENSE. The repository is gated. Agree to…
Independent publisher
shengjieXu
layce
GNN&BME&NLP
Models in Library0
Datasets in Library1
Models on Hugging Face—
Followers—
Compare the published evidence
This profile connects the 1 model and dataset record shown below to their upstream cards and sources. The library counts describe SAVRN's coverage, rather than this publisher's entire output.
Published tasks: Text classification.
Reported licenses: other. Open a record to review its specific license and conditions.
Compare model specifications, source dates, and hardware estimates on the individual records. Publisher identity verification is separate from a model's quality, deployment readiness, or suitability for your work.