# Judging Quality Across Languages: A Multilingual Approach to Pretraining Data Filtering with Language Models

arXiv: https://arxiv.org/abs/2505.22232 · HTML: https://mlsanity.com/paper/2505.22232/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 28 May 2025.

## Built on by (most cited first, 7 of 7)

- [propella-1: Multi-Property Document Annotation for LLM Data Curation at Scale](https://mlsanity.com/paper/2602.12414/) (2026)
- [Multilinguality at the Edge: Developing Language Models for the Global South](https://mlsanity.com/paper/2604.21637/) (2026)
- [Improving Romanian LLM Pretraining Data using Diversity and Quality Filtering](https://mlsanity.com/paper/2511.01090/) (2025)
- [Raising Bars, Not Parameters: LilMoo Compact Language Model for Hindi](https://mlsanity.com/paper/2603.03508/) (2026)
- [MameLoshnLM: Yiddish Language Model and Evaluation Benchmark](https://mlsanity.com/paper/2608.05850/) (2026)
- [Fine PT-PT Web: A High-Quality 41 Billion Tokens Data Collection of the European Portuguese Web](https://mlsanity.com/paper/2609.07699/) (2026)
- [Host-Aware Web Crawling for LLM Pretraining Data](https://mlsanity.com/paper/2610.13321/) (2026)

## Abstract (arXiv)

High-quality multilingual training data is essential for effectively pretraining large language models (LLMs). Yet, the availability of suitable open-source multilingual datasets remains limited. Existing state-of-the-art datasets mostly rely on heuristic filtering methods, restricting both their cross-lingual transferability and scalability. Here, we introduce JQL, a systematic approach that efficiently curates diverse and high-quality multilingual data at scale while significantly reducing computational demands. JQL distills LLMs' annotation capabilities into lightweight annotators based on pretrained multilingual embeddings. These models exhibit robust multilingual and cross-lingual performance, even for languages and scripts unseen during training. Evaluated empirically across 35 languages, the resulting annotation pipeline substantially outperforms current heuristic filtering methods like Fineweb2. JQL notably enhances downstream model training quality and increases data retention rates. Our research provides practical insights and valuable resources for multilingual data curation, raising the standards of multilingual dataset development.