Saved in:
Bibliographic Details
Main Authors: Arnett, Catherine, Hudspeth, Marisa, O'Connor, Brendan
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2507.06378
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918086906478592
author Arnett, Catherine
Hudspeth, Marisa
O'Connor, Brendan
author_facet Arnett, Catherine
Hudspeth, Marisa
O'Connor, Brendan
contents While tokenization is a key step in language modeling, with effects on model training and performance, it remains unclear how to effectively evaluate tokenizer quality. One proposed dimension of tokenizer quality is the extent to which tokenizers preserve linguistically meaningful subwords, aligning token boundaries with morphological boundaries within a word. We expand MorphScore (Arnett & Bergen, 2025), which previously covered 22 languages, to support a total of 70 languages. The updated MorphScore offers more flexibility in evaluation and addresses some of the limitations of the original version. We then correlate our alignment scores with downstream task performance for five pre-trained languages models on seven tasks, with at least one task in each of the languages in our sample. We find that morphological alignment does not explain very much variance in model performance, suggesting that morphological alignment alone does not measure dimensions of tokenization quality relevant to model performance.
format Preprint
id arxiv_https___arxiv_org_abs_2507_06378
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Evaluating Morphological Alignment of Tokenizers in 70 Languages
Arnett, Catherine
Hudspeth, Marisa
O'Connor, Brendan
Computation and Language
While tokenization is a key step in language modeling, with effects on model training and performance, it remains unclear how to effectively evaluate tokenizer quality. One proposed dimension of tokenizer quality is the extent to which tokenizers preserve linguistically meaningful subwords, aligning token boundaries with morphological boundaries within a word. We expand MorphScore (Arnett & Bergen, 2025), which previously covered 22 languages, to support a total of 70 languages. The updated MorphScore offers more flexibility in evaluation and addresses some of the limitations of the original version. We then correlate our alignment scores with downstream task performance for five pre-trained languages models on seven tasks, with at least one task in each of the languages in our sample. We find that morphological alignment does not explain very much variance in model performance, suggesting that morphological alignment alone does not measure dimensions of tokenization quality relevant to model performance.
title Evaluating Morphological Alignment of Tokenizers in 70 Languages
topic Computation and Language
url https://arxiv.org/abs/2507.06378