# An Empirical Comparison of Vocabulary Expansion and Initialization Approaches for Language Models

arXiv: https://arxiv.org/abs/2407.05841 · HTML: https://mlsanity.com/paper/2407.05841/

## Key facts

- Cited by 20 papers in the map (data through 14 Oct 2026).
- Builds on 30 of its 43 references that are in the map.
- First posted on arXiv 8 Jul 2024.

## Builds on (most cited first, 30 of 30)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [PaLM: Scaling Language Modeling with Pathways](https://mlsanity.com/paper/2204.02311/) (2022)
- [Unsupervised Cross-lingual Representation Learning at Scale](https://mlsanity.com/paper/1911.02116/) (2019)
- [Distributed Representations of Words and Phrases and their Compositionality](https://mlsanity.com/paper/1310.4546/) (2013)
- [Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing](https://mlsanity.com/paper/2107.13586/) (2021)
- [SentencePiece: A simple and language independent subword tokenizer and detokenizer for Neural Text Processing](https://mlsanity.com/paper/1808.06226/) (2018)
- [BLOOM: A 176B-Parameter Open-Access Multilingual Language Model](https://mlsanity.com/paper/2211.05100/) (2022)
- [Know What You Don't Know: Unanswerable Questions for SQuAD](https://mlsanity.com/paper/1806.03822/) (2018)
- [No Language Left Behind: Scaling Human-Centered Machine Translation](https://mlsanity.com/paper/2207.04672/) (2022)
- [Cross-lingual Language Model Pretraining](https://mlsanity.com/paper/1901.07291/) (2019)
- [XNLI: Evaluating Cross-lingual Sentence Representations](https://mlsanity.com/paper/1809.05053/) (2018)
- [How multilingual is Multilingual BERT?](https://mlsanity.com/paper/1906.01502/) (2019)
- [On the Cross-lingual Transferability of Monolingual Representations](https://mlsanity.com/paper/1910.11856/) (2019)
- [CamemBERT: a Tasty French Language Model](https://mlsanity.com/paper/1911.03894/) (2019)
- [AraBERT: Transformer-based Model for Arabic Language Understanding](https://mlsanity.com/paper/2003.00104/) (2020)
- [Efficient and Effective Text Encoding for Chinese LLaMA and Alpaca](https://mlsanity.com/paper/2304.08177/) (2023)
- [IndicTrans2: Towards High-Quality and Accessible Machine Translation Models for all 22 Scheduled Indian Languages](https://mlsanity.com/paper/2305.16307/) (2023)
- [A Monolingual Approach to Contextualized Word Embeddings for Mid-Resource Languages](https://mlsanity.com/paper/2006.06202/) (2020)
- [SeaLLMs -- Large Language Models for Southeast Asia](https://mlsanity.com/paper/2312.00738/) (2023)
- [Towards Leaving No Indic Language Behind: Building Monolingual Corpora, Benchmark and Models for Indic Languages](https://mlsanity.com/paper/2212.05409/) (2022)
- [WECHSEL: Effective initialization of subword embeddings for cross-lingual transfer of monolingual language models](https://mlsanity.com/paper/2112.06598/) (2021)
- [IndicLLMSuite: A Blueprint for Creating Pre-training and Fine-Tuning Datasets for Indian Languages](https://mlsanity.com/paper/2403.06350/) (2024)
- [Fast Vocabulary Transfer for Language Model Compression](https://mlsanity.com/paper/2402.09977/) (2024)
- [From English To Foreign Languages: Transferring Pre-trained Language Models](https://mlsanity.com/paper/2002.07306/) (2020)
- [OFA: A Framework of Initializing Unseen Subword Embeddings for Efficient Large-scale Multilingual Continued Pretraining](https://mlsanity.com/paper/2311.08849/) (2023)
- [Towards Continual Learning for Multilingual Machine Translation via Vocabulary Substitution](https://mlsanity.com/paper/2103.06799/) (2021)
- [InstructAlign: High-and-Low Resource Language Alignment via Continual Crosslingual Instruction Tuning](https://mlsanity.com/paper/2305.13627/) (2023)
- [Indian Language Wordnets and their Linkages with Princeton WordNet](https://mlsanity.com/paper/2201.02977/) (2022)

## Built on by (most cited first, 20 of 20)

- [TokAlign: Efficient Vocabulary Adaptation via Token Alignment](https://mlsanity.com/paper/2506.03523/) (2025)
- [Adapters for Altering LLM Vocabularies: What Languages Benefit the Most?](https://mlsanity.com/paper/2410.09644/) (2024)
- [The Vision Wormhole: Latent-Space Communication in Heterogeneous Multi-Agent Systems](https://mlsanity.com/paper/2602.15382/) (2026)
- [Continually Adding New Languages to Multilingual Language Models](https://mlsanity.com/paper/2509.11414/) (2025)
- [Adapting Chat Language Models Using Only Target Unlabeled Language Data](https://mlsanity.com/paper/2412.11704/) (2024)
- [Token Distillation: Attention-aware Input Embeddings For New Tokens](https://mlsanity.com/paper/2505.20133/) (2025)
- [RASMALAI: Resources for Adaptive Speech Modeling in Indian Languages with Accents and Intonations](https://mlsanity.com/paper/2505.18609/) (2025)
- [Mitigating Catastrophic Forgetting in Target Language Adaptation of LLMs via Source-Shielded Updates](https://mlsanity.com/paper/2512.04844/) (2025)
- [HYPEROFA: Expanding LLM Vocabulary to New Languages via Hypernetwork-Based Embedding Initialization](https://mlsanity.com/paper/2504.21018/) (2025)
- [Semantic Aware Linear Transfer by Recycling Pre-trained Language Models for Cross-lingual Transfer](https://mlsanity.com/paper/2505.10945/) (2025)
- [Defragmenting Language Models: An Interpretability-based Approach for Vocabulary Expansion](https://mlsanity.com/paper/2604.16656/) (2026)
- [Vocabulary Customization for Efficient Domain-Specific LLM Deployment](https://mlsanity.com/paper/2509.26124/) (2025)
- [TokAlign++: Advancing Vocabulary Adaptation via Better Token Alignment](https://mlsanity.com/paper/2605.13429/) (2026)
- [MultiHashFormer: Hash-based Generative Language Models](https://mlsanity.com/paper/2606.28057/) (2026)
- [In-Place Tokenizer Expansion for Pre-trained LLMs](https://mlsanity.com/paper/2607.15232/) (2026)
- [Prompt, Translate, Fine-Tune, Re-Initialize, or Instruction-Tune? Adapting LLMs for In-Context Learning in Low-Resource Languages](https://mlsanity.com/paper/2506.19187/) (2025)
- [LGSE: Lexically Grounded Subword Embedding Initialization for Low-Resource Language Adaptation](https://mlsanity.com/paper/2603.22629/) (2026)
- [Large Language Models as Amortized Pareto-Front Generators for Constrained Bi-Objective Convex Optimization](https://mlsanity.com/paper/2605.12106/) (2026)
- [Why Advanced Encoders Lag on Sparse Retrieval? The Answer and an Approach to Bridging Vocabulary Gaps](https://mlsanity.com/paper/2607.00004/) (2026)
- [Osprey: Target-agnostic Pre-training Makes Stronger Drafters in Speculative Decoding](https://mlsanity.com/paper/2609.09338/) (2026)

## Abstract (arXiv)

Language Models (LMs) excel in natural language processing tasks for English but show reduced performance in most other languages. This problem is commonly tackled by continually pre-training and fine-tuning these models for said languages. A significant issue in this process is the limited vocabulary coverage in the original model's tokenizer, leading to inadequate representation of new languages and necessitating an expansion of the tokenizer. The initialization of the embeddings corresponding to new vocabulary items presents a further challenge. Current strategies require cross-lingual embeddings and lack a solid theoretical foundation as well as comparisons with strong baselines. In this paper, we first establish theoretically that initializing within the convex hull of existing embeddings is a good initialization, followed by a novel but simple approach, Constrained Word2Vec (CW2V), which does not require cross-lingual embeddings. Our study evaluates different initialization methods for expanding RoBERTa and LLaMA 2 across four languages and five tasks. The results show that CW2V performs equally well or even better than more advanced techniques. Additionally, simpler approaches like multivariate initialization perform on par with these advanced methods indicating that efficient large-scale multilingual continued pretraining can be achieved even with simpler initialization methods. We release our code publicly (https://github.com/AI4Bharat/VocabAdaptation_LLM/tree/CW2V).