TokAlign: Efficient Vocabulary Adaptation via Token Alignment

• URL: http://arxiv.org/abs/2506.03523v1
• Date: Wed, 04 Jun 2025 03:15:57 GMT
• Title: TokAlign: Efficient Vocabulary Adaptation via Token Alignment
• Authors: Chong Li, Jiajun Zhang, Chengqing Zong,
• Abstract summary: Tokenization serves as a foundational step for Large Language Models (LLMs) to process text.<n>In new domains or languages, the inefficiency of the tokenizer will slow down the training and generation of LLM.<n>We propose an efficient method named TokAlign to replace the vocabulary of LLM from the token co-occurrences view.
• Score: 41.59130966729569
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Tokenization serves as a foundational step for Large Language Models (LLMs) to process text. In new domains or languages, the inefficiency of the tokenizer will slow down the training and generation of LLM. The mismatch in vocabulary also hinders deep knowledge transfer between LLMs like token-level distillation. To mitigate this gap, we propose an efficient method named TokAlign to replace the vocabulary of LLM from the token co-occurrences view, and further transfer the token-level knowledge between models. It first aligns the source vocabulary to the target one by learning a one-to-one mapping matrix for token IDs. Model parameters, including embeddings, are rearranged and progressively fine-tuned for the new vocabulary. Our method significantly improves multilingual text compression rates and vocabulary initialization for LLMs, decreasing the perplexity from 3.4$\text{e}^2$ of strong baseline methods to 1.2$\text{e}^2$ after initialization. Experimental results on models across multiple parameter scales demonstrate the effectiveness and generalization of TokAlign, which costs as few as 5k steps to restore the performance of the vanilla model. After unifying vocabularies between LLMs, token-level distillation can remarkably boost (+4.4% than sentence-level distillation) the base model, costing only 235M tokens.

Related papers

• Sampling from Your Language Model One Byte at a Time [82.71473348639489]
  Tokenization can introduce distortion into the model's generations, known as the Prompt Boundary Problem (PBP)<n>We present an inference-time method to convert any autore LM with a BPE tokenizer into a character-level or byte-level LM.<n>Our method efficiently solves the PBP and is also able to unify the vocabularies of language models with different tokenizers.
  arXiv  Detail & Related papers  (2025-06-17T02:37:04Z)
• Training-Free Tokenizer Transplantation via Orthogonal Matching Pursuit [45.18582668677648]
  We present a training-free method to transplant tokenizers in large language models.<n>We approximate each out-of-vocabulary token as a sparse linear combination of shared tokens.<n>We show that OMP achieves best zero-shot preservation of the base model's performance.
  arXiv  Detail & Related papers  (2025-06-07T00:51:27Z)
• Segment First or Comprehend First? Explore the Limit of Unsupervised Word Segmentation with Large Language Models [92.92512796044471]
  We propose a new framework to explore the limit of unsupervised word segmentation with Large Language Models (LLMs)<n>We employ current mainstream LLMs to perform word segmentation across multiple languages to assess LLMs' "comprehension"<n>We introduce a novel unsupervised method, termed LLACA, which enables the construction of a dynamic $n$-gram model that adjusts based on contextual information.
  arXiv  Detail & Related papers  (2025-05-26T07:48:15Z)
• Tokenization is Sensitive to Language Variation [14.568179478275255]
  Tokenizers split texts into smaller units and might behave differently for less common linguistic forms.<n>This might affect downstream LLM performance differently on two types of tasks.<n>We investigate how key algorithmic design choices impact downstream models' performances.
  arXiv  Detail & Related papers  (2025-02-21T09:58:54Z)
• Token Prepending: A Training-Free Approach for Eliciting Better Sentence Embeddings from LLMs [10.213016513358598]
  Token Prepending (TP) technique prepends each layer's decoded sentence embedding to the beginning of the sentence in the next layer's input.<n>TP technique is a plug-and-play and training-free technique, which means it can be seamlessly integrated with prompt-based sentence embedding methods.
  arXiv  Detail & Related papers  (2024-12-16T08:42:00Z)
• Retrofitting Large Language Models with Dynamic Tokenization [3.608780819053423]
  We propose retrofitting current language models with dynamic tokenization.<n>We merge frequent subword sequences in a batch, then apply a pre-trained embedding-prediction hypernetwork to compute the token embeddings on-the-fly.<n>We find that dynamic tokenization can mitigate the limitations of static tokenization by substantially improving inference speed and promoting fairness across languages.
  arXiv  Detail & Related papers  (2024-11-27T17:51:58Z)
• Cool-Fusion: Fuse Large Language Models without Training [73.17551121242602]
  emphCool-Fusion is a method that does not require any type of training like the ensemble approaches. emphCool-Fusion increases accuracy from three strong source LLMs by a significant 8%-17.8%.
  arXiv  Detail & Related papers  (2024-07-29T09:02:19Z)
• IntactKV: Improving Large Language Model Quantization by Keeping Pivot Tokens Intact [46.32830393597601]
  Large language models (LLMs) excel in natural language processing but demand intensive computation. This paper unveils a previously overlooked type of outliers in LLMs. We propose IntactKV to generate the KV cache of pivot tokens losslessly from the full-precision model.
  arXiv  Detail & Related papers  (2024-03-02T16:05:26Z)
• The Ups and Downs of Large Language Model Inference with Vocabulary Trimming by Language Heuristics [74.99898531299148]
  This research examines vocabulary trimming (VT) inspired by restricting embedding entries to the language of interest to bolster time and memory efficiency. We apply two languages to trim the full vocabulary - Unicode-based script filtering and corpus-based selection - to different language families and sizes. It is found that VT reduces the memory usage of small models by nearly 50% and has an upper bound of 25% improvement in generation speed.
  arXiv  Detail & Related papers  (2023-11-16T09:35:50Z)
• OFA: A Framework of Initializing Unseen Subword Embeddings for Efficient Large-scale Multilingual Continued Pretraining [49.213120730582354]
  Instead of pretraining multilingual language models from scratch, a more efficient method is to adapt existing pretrained language models (PLMs) to new languages via vocabulary extension and continued pretraining. We propose a novel framework: $textbfO$ne $textbfF$or $textbfA$ll, which wisely initializes the embeddings of unseen subwords and thus can adapt a PLM to multiple languages efficiently and effectively.
  arXiv  Detail & Related papers  (2023-11-15T10:40:45Z)
• Translate to Disambiguate: Zero-shot Multilingual Word Sense Disambiguation with Pretrained Language Models [67.19567060894563]
  Pretrained Language Models (PLMs) learn rich cross-lingual knowledge and can be finetuned to perform well on diverse tasks. We present a new study investigating how well PLMs capture cross-lingual word sense with Contextual Word-Level Translation (C-WLT) We find that as the model size increases, PLMs encode more cross-lingual word sense knowledge and better use context to improve WLT performance.
  arXiv  Detail & Related papers  (2023-04-26T19:55:52Z)
• Byte Pair Encoding is Suboptimal for Language Model Pretraining [49.30780227162387]
  We analyze differences between unigram LM tokenization and byte-pair encoding (BPE) We find that the unigram LM tokenization method matches or outperforms BPE across downstream tasks and two languages. We hope that developers of future pretrained LMs will consider adopting the unigram LM method over the more prevalent BPE.
  arXiv  Detail & Related papers  (2020-04-07T21:21:06Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.