# Revisiting Multilingual Data Mixtures in Language Model Pretraining

arXiv: https://arxiv.org/abs/2510.25947 · HTML: https://mlsanity.com/paper/2510.25947/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- Builds on 49 of its 60 references that are in the map.
- First posted on arXiv 29 Oct 2025.

## Builds on (most cited first, 30 of 49)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [LLaMA: Open and Efficient Foundation Language Models](https://mlsanity.com/paper/2302.13971/) (2023)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [Unsupervised Cross-lingual Representation Learning at Scale](https://mlsanity.com/paper/1911.02116/) (2019)
- [CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledge](https://mlsanity.com/paper/1811.00937/) (2018)
- [mT5: A massively multilingual pre-trained text-to-text transformer](https://mlsanity.com/paper/2010.11934/) (2020)
- [DoReMi: Optimizing Data Mixtures Speeds Up Language Model Pretraining](https://mlsanity.com/paper/2305.10429/) (2023)
- [XCOPA: A Multilingual Dataset for Causal Commonsense Reasoning](https://mlsanity.com/paper/2005.00333/) (2020)
- [Data Selection for Language Models via Importance Resampling](https://mlsanity.com/paper/2302.03169/) (2023)
- [The Belebele Benchmark: a Parallel Reading Comprehension Dataset in 122 Language Variants](https://mlsanity.com/paper/2308.16884/) (2023)
- [Aya Model: An Instruction Finetuned Open-Access Multilingual Language Model](https://mlsanity.com/paper/2402.07827/) (2024)
- [Okapi: Instruction-tuned Large Language Models in Multiple Languages with Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2307.16039/) (2023)
- [RegMix: Data Mixture as Regression for Language Model Pre-training](https://mlsanity.com/paper/2407.01492/) (2024)
- [Scaling Laws and Compute-Optimal Training Beyond Fixed Training Durations](https://mlsanity.com/paper/2405.18392/) (2024)
- [UniMax: Fairer and more Effective Language Sampling for Large-Scale Multilingual Pretraining](https://mlsanity.com/paper/2304.09151/) (2023)
- [Lifting the Curse of Multilinguality by Pre-training Modular Transformers](https://mlsanity.com/paper/2205.06266/) (2022)
- [FineWeb2: One Pipeline to Scale Them All -- Adapting Pre-Training Data Processing to Every Language](https://mlsanity.com/paper/2506.20920/) (2025)
- [M3Exam: A Multilingual, Multimodal, Multilevel Benchmark for Examining Large Language Models](https://mlsanity.com/paper/2306.05179/) (2023)
- [DoGE: Domain Reweighting with Generalization Estimation](https://mlsanity.com/paper/2310.15393/) (2023)
- [INCLUDE: Evaluating Multilingual Language Understanding with Regional Knowledge](https://mlsanity.com/paper/2411.19799/) (2024)
- [Efficient Online Data Mixing For Language Model Pre-Training](https://mlsanity.com/paper/2312.02406/) (2023)
- [It's All in the Heads: Using Attention Heads as a Baseline for Cross-Lingual Transfer in Commonsense Reasoning](https://mlsanity.com/paper/2106.12066/) (2021)
- [Common Sense Beyond English: Evaluating and Improving Multilingual Language Models for Commonsense Reasoning](https://mlsanity.com/paper/2106.06937/) (2021)
- [Meta CLIP 2: A Worldwide Scaling Recipe](https://mlsanity.com/paper/2507.22062/) (2025)
- [EXAMS: A Multi-Subject High School Examinations Dataset for Cross-Lingual and Multilingual Question Answering](https://mlsanity.com/paper/2011.03080/) (2020)
- [PLUG: Leveraging Pivot Language in Cross-Lingual Instruction Tuning](https://mlsanity.com/paper/2311.08711/) (2023)

## Built on by (most cited first, 10 of 10)

- [E-RayZer: Self-supervised 3D Reconstruction as Spatial Visual Pre-training](https://mlsanity.com/paper/2512.10950/) (2025)
- [Cross-Lingual Sentiment Misalignment: Auditing Multilingual Language Models for Inversion Risk, Dialectal Representation, and Affective Stability](https://mlsanity.com/paper/2602.17469/) (2026)
- ["UberWeb: Insights from Multilingual Curation for a 20-Trillion-Token Dataset](https://mlsanity.com/paper/2602.15210/) (2026)
- [Leveraging Routing Dynamics in Mixture-of-Experts Models for Efficient Language Adaptation](https://mlsanity.com/paper/2605.29714/) (2026)
- [On the Limits of Model Merging for Multilinguality in Pre-Training](https://mlsanity.com/paper/2605.25846/) (2026)
- [Speculative Decoding and the Curse of Multilinguality](https://mlsanity.com/paper/2605.30580/) (2026)
- [Soft Token Alignment for Cross-Lingual Reasoning](https://mlsanity.com/paper/2606.26466/) (2026)
- [LLM Scheming Inversely Scales with Pretraining Language Coverage](https://mlsanity.com/paper/2607.24769/) (2026)
- [There is No Theoretical Curse of Multilinguality For Embedding Space Structure](https://mlsanity.com/paper/2608.17088/) (2026)
- [Lingua Franca or Probing Artifact? Rethinking Latent Language in Multilingual LLMs](https://mlsanity.com/paper/2609.00155/) (2026)

## Abstract (arXiv)

The impact of different multilingual data mixtures in pretraining large language models (LLMs) has been a topic of ongoing debate, often raising concerns about potential trade-offs between language coverage and model performance (i.e., the curse of multilinguality). In this work, we investigate these assumptions by training 1.1B and 3B parameter LLMs on diverse multilingual corpora, varying the number of languages from 25 to 400. Our study challenges common beliefs surrounding multilingual training. First, we find that combining English and multilingual data does not necessarily degrade the in-language performance of either group, provided that languages have a sufficient number of tokens included in the pretraining corpus. Second, we observe that using English as a pivot language (i.e., a high-resource language that serves as a catalyst for multilingual generalization) yields benefits across language families, and contrary to expectations, selecting a pivot language from within a specific family does not consistently improve performance for languages within that family. Lastly, we do not observe a significant "curse of multilinguality" as the number of training languages increases in models at this scale. Our findings suggest that multilingual data, when balanced appropriately, can enhance language model capabilities without compromising performance, even in low-resource settings