XLM-R, pretrained on 100 languages with 2TB of CommonCrawl data, improves average XNLI accuracy by 14.6 points and MLQA F1 by 13 points over mBERT while matching strong monolingual models on GLUE.
arXiv preprint arXiv:1910.07475 , year=
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
representative citing papers
SHIFT mitigates language bias in MLIR by subtracting estimated relative language vectors from document embeddings during indexing using parallel translation pairs.
A tutorial synthesizing foundations, recent models such as PALO and Maya, and low-cost methods for tri-modal multilingual AI in resource-constrained settings.
citing papers explorer
-
Unsupervised Cross-lingual Representation Learning at Scale
XLM-R, pretrained on 100 languages with 2TB of CommonCrawl data, improves average XNLI accuracy by 14.6 points and MLQA F1 by 13 points over mBERT while matching strong monolingual models on GLUE.
-
SHIFT: Semantic Harmonization via Index-side Feature Transformation for Multilingual Information Retrieval
SHIFT mitigates language bias in MLIR by subtracting estimated relative language vectors from document embeddings during indexing using parallel translation pairs.
-
Multilingual and Multimodal LLMs in the Wild: Building for Low-Resource Languages
A tutorial synthesizing foundations, recent models such as PALO and Maya, and low-cost methods for tri-modal multilingual AI in resource-constrained settings.