UnihanLM: Coarse-to-Fine Chinese-Japanese Language Model Pretraining with the Unihan Database.

Canwen Xu, Tao Ge, Chenliang Li, Furu Wei

Browse the full IJCNLP paper archive.