Assessing the Role of Data Quality in Training Bilingual Language Models (2025.findings-emnlp)
Copied to clipboard
| Challenge: | a recent study shows that adding more languages can degrade performance for some languages while improving others. |
| Approach: | They propose a data filtering strategy to select high-quality bilingual training data with only high quality English data. |
| Outcome: | The proposed approach improves bilingual model performance by 2–4% and reduces bilingual models performance gaps to 1%. |
Similar Papers
Judging Quality Across Languages: A Multilingual Approach to Pretraining Data Filtering with Language Models (2025.emnlp-main)
Copied to clipboard
Mehdi Ali, Manuel Brack, Max Lübbering, Elias Wendt, Abbas Goher Khan, Richard Rutmann, Alex Jude, Maurice Kraus, Alexander Arno Weber, Felix Stollenwerk, David Kaczér, Florian Mai, Lucie Flek, Rafet Sifa, Nicolas Flores-Herr, Joachim Koehler, Patrick Schramowski, Michael Fromm, Kristian Kersting
| Challenge: | Existing open-source multilingual datasets rely on heuristic filtering methods restricting both their cross-lingual transferability and scalability. |
| Approach: | They propose a systematic approach that curates diverse and high-quality multilingual data at scale while significantly reducing computational demands. |
| Outcome: | Evaluated empirically across 35 languages, the proposed approach outperforms current heuristic filtering methods like Fineweb2 and improves model training quality and retention rates. |
When Is Multilinguality a Curse? Language Modeling for 250 High- and Low-Resource Languages (2024.emnlp-main)
Copied to clipboard
| Challenge: | Multilingual language models are widely used to extend NLP systems to low-resource languages. |
| Approach: | They pre-train over 10,000 monolingual and multilingual language models for over 250 languages including multiple language families that are under-studied in NLP. |
| Outcome: | The results show that adding multilingual data improves low-resource language modeling performance, similar to increasing low-source dataset sizes by up to 33%. |
The Role of Mixed-Language Documents for Multilingual Large Language Model Pretraining (2026.acl-long)
Copied to clipboard
Jiandong Shao, Raphael Tang, Crystina Zhang, Karin Sevegnani, Pontus Stenetorp, Jianfei Yang, Yao Lu
| Challenge: | Existing research suggests that multilingual large language models can achieve impressive cross-lingual understanding despite largely monolingual pretraining. |
| Approach: | They compare a monolingual-only corpus with a standard web corpus that removes all multilingual documents and then retrain the models from scratch under controlled conditions. |
| Outcome: | The results show that removing bilingual data causes translation performance to drop 56% in BLEU, whereas code-switching contributes minimally. |
Language Contamination Helps Explains the Cross-lingual Capabilities of English Pretrained Models (2022.emnlp-main)
Copied to clipboard
| Challenge: | a large number of pretraining corpora are not publicly available, and it is unclear how much foreign language data exists in monolingual models. |
| Approach: | They propose to use English pretraining corpora to analyze their language composition . they find that even when less than 1% of data is not English, it facilitates cross-lingual transfer . |
| Outcome: | The proposed model is not truly monolingual when pretrained at scale, the authors show . they show that even when less than 1% of data is not English, it facilitates cross-lingual transfer . |
Training Bilingual LMs with Data Constraints in the Targeted Language (2025.findings-acl)
Copied to clipboard
| Challenge: | a large number of languages have insufficient data for pretraining, but most non-English models are trained on scrapes of the web. |
| Approach: | They propose to use data from an auxiliary language to boost model performance . they quantify the performance gap between training with data in a data-rich auxiliary and training in the target language . |
| Outcome: | The proposed method boosts model performance in a target language with insufficient data . it also explores the benefits of translation systems and the limitations of model scaling when data is limited. |
Is It Good Data for Multilingual Instruction Tuning or Just Bad Multilingual Evaluation for Large Language Models? (2024.emnlp-main)
Copied to clipboard
| Challenge: | Existing practices of fine-tuning and evaluating multilingual large language models may not align with this objective due to a heavy reliance on translation. |
| Approach: | They propose to use translated or native instruction data to fine-tune multilingual large language models. |
| Outcome: | The proposed model can be fine tuned and evaluated in multilingual large language models . the results show that native or translated data can be used to compare model performance . |
How Can Synthetic Data Improve Multilingual Language Model Pretraining? A Data Quality Perspective (2026.acl-long)
Copied to clipboard
| Challenge: | Low-resource languages are a long-tail problem for multilingual LLMs due to limited high-quality training data. |
| Approach: | They propose a method that translates high-quality, knowledge-rich English data into low-resource languages . they propose SynRank, which leverages synthetic data as positive samples to train a classifier . |
| Outcome: | The proposed method matches handcrafted rule-based filtering by human experts and significantly improves knowledge-intensive tasks with less data. |
Comparing Biases and the Impact of Multilingual Training across Multiple Languages (2023.emnlp-main)
Copied to clipboard
Sharon Levy, Neha John, Ling Liu, Yogarshi Vyas, Jie Ma, Yoshinari Fujinuma, Miguel Ballesteros, Vittorio Castelli, Dan Roth
| Challenge: | Currently, studies on bias and fairness in natural language processing focus on a single language and/or across few attributes (e.g. gender, race). However, biases can manifest differently across languages for individual attributes. |
| Approach: | They adapt existing sentiment bias templates in English to Italian, Chinese, Hebrew, and Spanish for race, religion, nationality, and gender. |
| Outcome: | The proposed model favors groups that are dominant in each language's culture, indicating bias amplification, after multilingual finetuning. |
Expanding Pretrained Models to Thousands More Languages via Lexicon-based Adaptation (2022.acl-long)
Copied to clipboard
| Challenge: | Recent studies have found that the performance of multilingual pretrained models is highly dependent on the availability of monolingual or parallel text in a target language. |
| Approach: | They propose to use bilingual lexicons to synthesize textual or labeled data and combine it with monolingual or parallel text when available. |
| Outcome: | The proposed methods improve performance for 19 under-represented languages with and without extra monolingual text. |
An Exploratory Study on Multilingual Quality Estimation (2020.aacl-main)
Copied to clipboard
Shuo Sun, Marina Fomicheva, Frédéric Blain, Vishrav Chaudhary, Ahmed El-Kishky, Adithya Renduchintala, Francisco Guzmán, Lucia Specia
| Challenge: | Existing approaches to predict the quality of machine translation use language-specific models, but they lack labelled data for each language pair. |
| Approach: | They propose to use scores from translation models to estimate quality of machine translations by predicting the quality of a translation at test time. |
| Outcome: | The proposed models outperform single-language models in less balanced quality label distributions and low-resource settings. |