M2M-100: mined many-to-many translation beyond English pivots

What changed

Combined mined parallel data across non-English language directions with shared and language-specific capacity studies, releasing a many-to-many translation model and reproducible data/evaluation workflows.

Scientific scope

Combined mined parallel data across non-English language directions with shared and language-specific capacity studies, releasing a many-to-many translation model and reproducible data/evaluation workflows. Direct non-English translation does not imply supervised data for every possible pair. Inherits LASER/FAISS mining and Transformer/SentencePiece components. Language-specific normalization affects reported BLEU; capability and quality are not uniform across all100languages.

Evidence

The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Direct non-English translation does not imply supervised data for every possible pair. Inherits LASER/FAISS mining and Transformer/SentencePiece components. Language-specific normalization affects reported BLEU; capability and quality are not uniform across all100languages.

Contributors

• Source-listed authors and technical contributors · SCIENCE: Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.

• Beyond English-Centric Multilingual Machine Translation

Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.

Full scientific record · Browse contribution grades · Request a correction