Modelling the Morphology of Verbal Paradigms: A Case Study in the Tokenization of Turkish and Hebrew
Not on the map. Why?
Abstract (arXiv)
We investigate how transformer models represent complex verb paradigms in Turkish and Modern Hebrew, concentrating on how tokenization strategies shape this ability. Using the Blackbird Language Matrices task on natural data, we show that for Turkish -- with its transparent morphological markers -- both monolingual and multilingual models succeed, either when tokenization is atomic or when it breaks words into small subword units. For Hebrew, instead, monolingual and multilingual models diverge. A multilingual model using character-level tokenization fails to capture the language non-concatenative morphology, but a monolingual model with morpheme-aware segmentation performs well. Performance improves on more synthetic datasets, in all models.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 25 of its 60 references that are in the map. First posted on arXiv 5 Feb 2026.
Builds on
25 of 60 references.
References · 25
- 2015Neural Machine Translation of Rare Words with Subword Unitscited by 2,898
- 2019BLiMP: The Benchmark of Linguistic Minimal Pairs for Englishcited by 553
- 2021CANINE: Pre-training an Efficient Tokenization-Free Encoder for Language Representationcited by 229
- 2023Do All Languages Cost the Same? Tokenization in the Era of Commercial Language Modelscited by 186
- 2020When Being Unseen from mBERT is just the Beginning: Handling New Languages With Multilingual Language Modelscited by 131
- 2020Syntactic Structure from Deep Learningcited by 99
- 2025MultiBLiMP 1.0: A Massively Multilingual Benchmark of Linguistic Minimal Pairscited by 65
- 2022Impact of Tokenization on Language Models: An Analysis for Turkishcited by 62
- 2025Synthetic Data Generation Using Large Language Models: Advances in Text and Codecited by 61
- 2019Investigating BERT's Knowledge of Language: Five Analysis Methods with NPIscited by 58
- 2024Toward a Theory of Tokenization in LLMscited by 28
- 2018How Much Does Tokenization Affect Neural Machine Translation?cited by 26
13 more
- 2018Verb Argument Structure Alternations in Word and Sentence Embeddingscited by 25
- 2017Linguistically Motivated Vocabulary Reduction for Neural Machine Translation from Turkish to Englishcited by 23
- 2025MorphBPE: A Morpho-Aware Tokenizer Bridging Linguistic Complexity for Efficient LLM Training Across Morphologiescited by 23
- 2023Bayes in the age of intelligent machinescited by 22
- 2024Tokenization and Morphology in Multilingual Language Models: A Comparative Analysis of mT5 and ByT5cited by 14
- 2023How Abstract Is Linguistic Generalization in Large Language Models? Experiments with Argument Structurecited by 13
- 2023Blackbird language matrices (BLM), a new task for rule-like generalization in neural networks: Motivations and Formal Specificationscited by 9
- 2024Exploring syntactic information in sentence embeddings through multilingual subject-verb agreementcited by 6
- 2024Exploring Italian sentence embeddings properties through multi-taskingcited by 6
- 2023Explicit Morphological Knowledge Improves Pre-training of Language Models for Hebrewcited by 5
- 2022Blackbird's language matrices (BLMs): a new benchmark to investigate disentangled generalisation in neural networkscited by 4
- 2025Does Training on Synthetic Data Make Models Less Robust?cited by 3
- 2025Challenging the Abilities of Large Language Models in Italian: a Community Initiativecited by 1
Built on by
Citing papers per year: 2026: 2.
2 papers.