Impact of Task Diversity in Intermediate-Task Training on Zero-Shot Cross-Lingual Transfer Performance

Multilingual pre-trained models have achieved remarkable performance on cross-lingual transfer learning. Some multilingual models such as mBERT, have been pre-trained on unlabeled corpora, therefore the embeddings of different languages in the models may not be aligned very well. In this paper, we aim to improve the zero-shot cross-lingual transfer performance by proposing a pre-training task named Word-Exchange Aligning Model (WEAM), which uses the statistical alignment information as the prior knowledge to guide cross-lingual word prediction. We evaluate our model on multilingual machine rea

Research goal: What is the impact of task diversity in intermediate-task training (e.g., NLI, QA, NER) on zero-shot cross-lingual transfer performance across different model scales (e.g., 110M vs. 750M parameters) when evaluated on MLQA and XQuAD benchmarks?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 8.1/10.