Repository for the article:
- Gutierrez-Vasques Ximena, Bentz Christian, Sozinova Olga and Samardzic Tanja. "From characters to words: the turning point of BPE merges”. European Chapter of the Association for Computational Linguistics, Long Papers. 2021
The folder Correlations/ contains a summary of the measures reported in the article.
The folder Detailed/ contains scripts, corpus and detailed results per each language.
The folder Media/ contains slides and a poster of the article.