Impact of Multidomain Intermediate Tasks on XLM-R Zero-Shot Cross-Lingual Transfer in PAWS-X

Pre-trained multilingual language encoders, such as multilingual BERT and XLM-R, show great potential for zero-shot cross-lingual transfer. However, these multilingual encoders do not precisely align words and phrases across languages. Especially, learning alignments in the multilingual embedding space usually requires sentence-level or word-level parallel corpora, which are expensive to be obtained for low-resource languages. An alternative is to make the multilingual encoders more robust; when fine-tuning the encoder using downstream task, we train the encoder to tolerate noise in the contex

Research goal: What is the impact of combining multiple intermediate language-understanding tasks from different domains (e.g., legal + medical) on the zero-shot cross-lingual transfer performance of XLM-R on the PAWS-X benchmark, compared to using a single intermediate task?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 7.8/10.