Cross-lingual Transfer of Monolingual Models
This addresses cross-lingual learning for NLP practitioners by enabling monolingual models to transfer knowledge, though it is incremental as it builds on existing domain adaptation techniques.
The paper tackles cross-lingual transfer for monolingual models using domain adaptation, showing that transferred models outperform native English models on GLUE benchmarks, with semantic information retained from the source language and syntactic information learned during transfer.
Recent studies in zero-shot cross-lingual learning using multilingual models have falsified the previous hypothesis that shared vocabulary and joint pre-training are the keys to cross-lingual generalization. Inspired by this advancement, we introduce a cross-lingual transfer method for monolingual models based on domain adaptation. We study the effects of such transfer from four different languages to English. Our experimental results on GLUE show that the transferred models outperform the native English model independently of the source language. After probing the English linguistic knowledge encoded in the representations before and after transfer, we find that semantic information is retained from the source language, while syntactic information is learned during transfer. Additionally, the results of evaluating the transferred models in source language tasks reveal that their performance in the source domain deteriorates after transfer.