English
Related papers

Related papers: BPE Gets Picky: Efficient Vocabulary Refinement Du…

200 papers

Modern language models still rely on fixed, pre-defined subword tokenizations. Once a tokenizer is trained, the LM can only operate at this fixed level of granularity, which often leads to brittle and counterintuitive behaviors even in…

Computation and Language · Computer Science 2026-03-05 Chunyuan Deng , Sanket Lokegaonkar , Colin Lockard , Besnik Fetahu , Nasser Zalmout , Xian Li

Transformers achieve unrivalled performance in modelling language, but remain inefficient in terms of memory and time complexity. A possible remedy is to reduce the sequence length in the intermediate layers by pooling fixed-length segments…

Computation and Language · Computer Science 2023-10-25 Piotr Nawrot , Jan Chorowski , Adrian Łańcucki , Edoardo M. Ponti

Tokenization is a crucial step in information retrieval, especially for lexical matching algorithms, where the quality of indexable tokens directly impacts the effectiveness of a retrieval system. Since different languages have unique…

Computation and Language · Computer Science 2022-10-12 Odunayo Ogundepo , Xinyu Zhang , Jimmy Lin

Computing next-token likelihood ratios between two language models (LMs) is a standard task in training paradigms such as knowledge distillation. Since this requires both models to share the same probability space, it becomes challenging…

Computation and Language · Computer Science 2026-05-07 Buu Phan , Ashish Khisti , Karen Ullrich

Cross-lingual vocabulary transfer plays a promising role in adapting pre-trained language models to new languages, including low-resource languages. Existing approaches that utilize monolingual or parallel corpora face challenges when…

Computation and Language · Computer Science 2025-06-03 Haruki Sakajo , Yusuke Ide , Justin Vasselli , Yusuke Sakai , Yingtao Tian , Hidetaka Kamigaito , Taro Watanabe

This paper presents a comprehensive examination of the impact of tokenization strategies and vocabulary sizes on the performance of Arabic language models in downstream natural language processing tasks. Our investigation focused on the…

Computation and Language · Computer Science 2024-09-23 Mohamed Taher Alrefaie , Nour Eldin Morsy , Nada Samir

Subword tokenization is a key design choice for modern language models, including large language models (LLMs), with byte- and character-level BPE serving as a widely used baseline. Standard BPE selects merges by raw pair frequency, which…

Computation and Language · Computer Science 2026-03-23 Azam Nouri

Adapting multilingual language models to specific languages can enhance both their efficiency and performance. In this study, we explore how modifying the vocabulary of a multilingual encoder model to better suit the Estonian language…

Computation and Language · Computer Science 2025-01-07 Aleksei Dorkin , Taido Purason , Kairit Sirts

Byte-pair encoding (BPE) is a ubiquitous algorithm in the subword tokenization process of language models as it provides multiple benefits. However, this process is solely based on pre-training data statistics, making it hard for the…

Computation and Language · Computer Science 2021-09-27 Gustavo Aguilar , Bryan McCann , Tong Niu , Nazneen Rajani , Nitish Keskar , Thamar Solorio

We introduce three simple randomized variants of byte pair encoding (BPE) and explore whether randomizing the selection of merge operations substantially affects a downstream machine translation task. We focus on translation into…

Computation and Language · Computer Science 2023-05-05 Jonne Sälevä , Constantine Lignos

Language models often pre-train on large unsupervised text corpora, then fine-tune on additional task-specific data. However, typical fine-tuning schemes do not prioritize the examples that they tune on. We show that, if you can prioritize…

Computation and Language · Computer Science 2023-05-12 Ian Osband , Seyed Mohammad Asghari , Benjamin Van Roy , Nat McAleese , John Aslanides , Geoffrey Irving

Tokenization is the act of breaking down text into smaller parts, or tokens, that are easier for machines to process. This is a key phase in machine translation (MT) models. Subword tokenization enhances this process by breaking down words…

Computation and Language · Computer Science 2025-05-23 Sudhansu Bala Das , Samujjal Choudhury , Tapas Kumar Mishra , Bidyut Kr. Patra

Byte-Pair Encoding (BPE) is an algorithm commonly used in Natural Language Processing to build a vocabulary of subwords, which has been recently applied to symbolic music. Given that symbolic music can differ significantly from text,…

Information Retrieval · Computer Science 2024-10-03 Dinh-Viet-Toan Le , Louis Bigo , Mikaela Keller

The pretraining data of today's strongest language models is opaque; in particular, little is known about the proportions of various domains or languages represented. In this work, we tackle a task which we call data mixture inference,…

Computation and Language · Computer Science 2024-12-03 Jonathan Hayase , Alisa Liu , Yejin Choi , Sewoong Oh , Noah A. Smith

Traditionally, NLP performance improvement has been focused on improving models and increasing the number of model parameters. NLP vocabulary construction has remained focused on maximizing the number of words represented through subword…

Computation and Language · Computer Science 2023-04-26 Sandeep Mehta , Darpan Shah , Ravindra Kulkarni , Cornelia Caragea

Byte Pair Encoding (BPE) tokenizers, widely used in Large Language Models, face challenges in multilingual settings, including penalization of non-Western scripts and the creation of tokens with partial UTF-8 sequences. Pretokenization,…

Computation and Language · Computer Science 2025-06-02 Sander Land , Catherine Arnett

Tokenization is the first step in modern neural language model pipelines where an input text is converted to a sequence of subword tokens. We introduce from first principles a finite-state transduction framework which can efficiently encode…

Computation and Language · Computer Science 2024-10-22 Marco Cognetta , Naoaki Okazaki

In recent years, language models have become increasingly larger and more complex. However, the input representations for these models continue to rely on simple and greedy subword tokenization methods. In this paper, we propose a novel…

Computation and Language · Computer Science 2023-06-14 David Samuel , Lilja Øvrelid

Tokenization is an important text preprocessing step to prepare input tokens for deep language models. WordPiece and BPE are de facto methods employed by important models, such as BERT and GPT. However, the impact of tokenization can be…

Computation and Language · Computer Science 2023-03-28 Cagri Toraman , Eyup Halit Yilmaz , Furkan Şahinuç , Oguzhan Ozcelik

Language prediction is constrained by informational entropy intrinsic to language, such that there exists a limit to how accurate any language model can become and equivalently a lower bound to language compression. The most efficient…

Computation and Language · Computer Science 2025-11-14 Benjamin L. Badger , Matthew Neligeorge