arXiv:2607. 22662v1 Announce Type: new Abstract: Open-web corpora curated via highly selective filters, such as FineWeb-Edu and DCLM, constitute the core of LLM pretraining data and have significantly advanced LLM performance.
By Peiguang Li, Yongwei Zhou, Juncheng Diao, Yuchun Fan, Jian Yang, Jianxiao Yang, Zhongda Su, Shuguang Jiao, Xiao Wei, Zhiye Zou, Gan Dong, Zhizhao Zeng, Rongxiang Weng, Jingang Wang, Xunliang Cai
arXiv:2607.00890v2 Announce Type: replace
Abstract: Open web-scale pre-training corpora remain concentrated in English, limiting multilingual LLM development. We introduce MultiSynt/MT, an open synth...
By Maximilian Idahl, J\"org Tiedemann, Sampo Pyysalo, David Salinas, Tomasz Galica, Shenbin Qian, Tudor Nicolae Mateiu, Zihao Li, Anna Lokrantz, Fedor Vitiugin, Andr\'e F. T. Martins, Jenna Kanerva, Filip Ginter, Matthias Lindemann, Tim Isbister, Birger Moell, Jonas Lindh, Jan Haji\v{c}, Jenia Jitsev, Andrey Kutuzov, Stephan Oepen, Gema Ram\'irez-S\'anchez
Open web-scale pre-training corpora remain concentrated in English, limiting multilingual LLM development. We introduce MultiSynt/MT, an open synthetic parallel corpus with approximately 4.
Large-scale pretraining corpora contain substantial duplicate content. Although document-level deduplication is widely used, removing subdocument-level redundancy remains challenging.
The paper introduces the problem of cross‑lingual loopholes in large language model (LLM) unlearning, where forgetting a fact in one language can leave it accessible in others. It presents a new 174‑language benchmark, the Cross‑Lingual Unlearning Tensor, and proposes COVER, a method that selects a subset of source languages to maximize unlearning coverage under a language budget. Experiments show COVER reduces residual knowledge by 7.8–27.3% compared to uniform selection and works on both synthetic and real low‑resource news data.
By Tyler Skow, Shravan Chaudhari, Rama Chellappa, Abhay Yadav
TransClean introduces a benchmark for identifying and removing translation noise—unwanted text such as language labels, explanations, or bilingual repetitions—from large language model (LLM) outputs. The authors analyzed 790,000 translations from 12 LLMs across 22 language pairs, cataloguing 12 common noise patterns and creating 9,900 noisy‑clean pairs (8,800 synthetic, 1,100 authentic). They evaluated two extraction methods—a span‑based approach using quality estimation models and an LLM‑prompted method—demonstrating the first systematic framework to assess and improve translation cleanliness.
By Shenbin Qian, Yves Scherrer