Tokenizer Fertility and Zero-Shot Performance of Foundation Models on Ukrainian Legal Text: A Comparative Study
Volodymyr Ovcharov · 2026
Read Tokenizer Fertility and Zero-Shot Performance of Foundation Models on Ukrainian Legal Text: A Comparative Study by Volodymyr Ovcharov (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-14890.txt.
Description
Ovcharov benchmarks tokenizer fertility and zero-shot performance on Ukrainian legal text. Fertility varies 1.6-fold across models, yet this cost-critical dimension is missing from model selection. Seven models from five providers are tested on two hundred and seventy-three validated court decisions from Ukraine's state registry. Qwen 3 models consume sixty percent more tokens than Llama-family models. NVIDIA Nemotron Super 3 scores highest, beating a model with 5.6 times more parameters at one third the API cost. Few-shot prompting cuts performance by up to twenty-six percentage points.
Catalog record
| CREATOR | Volodymyr Ovcharov |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 70266 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.14890