AWhat changed
Studied shared multilingual causal/masked language objectives and parallel-sentence translation language modeling; XLM-R extended masked pretraining with larger filtered web corpora, vocabulary and capacity tradeoff analysis.
Scientific scope
Studied shared multilingual causal/masked language objectives and parallel-sentence translation language modeling; XLM-R extended masked pretraining with larger filtered web corpora, vocabulary and capacity tradeoff analysis. XLM's parallel-data objective is supervised; XLM-R's monolingual pretraining does not imply all evaluations are unsupervised. Improvements inherit BERT/RoBERTa and depend on sampling, capacity and resource distribution. Original v1 results differ from later revised abstracts, so no cross-version numerical claim is imported.
Evidence
The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. XLM's parallel-data objective is supervised; XLM-R's monolingual pretraining does not imply all evaluations are unsupervised. Improvements inherit BERT/RoBERTa and depend on sampling, capacity and resource distribution. Original v1 results differ from later revised abstracts, so no cross-version numerical claim is imported.
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.
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