Author

G. S. Altintas

1 paper indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

#natural language process... Preprint Oct 2026

Language-Specific Effects of Tokenizer Choice in Multilingual Language Models

It is found that tokenizer choice matters more for languages with less language-model training data: across the 54 tokenizers, the standard deviation of a language's bits-per-byte (BPB) increases as its model training-data share decreases.

Clara Meister, G. S. Altintas, Antoine Bosselut · 0 citations