Cross-lingual Knowledge Distillation for Zero-shot XOR-TyDi QA Performance

Benefiting from transformer-based pre-trained language models, neural ranking models have made significant progress. More recently, the advent of multilingual pre-trained language models provides great support for designing neural cross-lingual retrieval models. However, due to unbalanced pre-training data in different languages, multilingual language models have already shown a performance gap between high and low-resource languages in many downstream tasks. And cross-lingual retrieval models built on such pre-trained models can inherit language bias, leading to suboptimal result for low-reso

Research goal: To what extent does cross-lingual knowledge distillation from high-resource languages to low-resource languages improve zero-shot performance on XOR-TyDi QA, measured by exact match accuracy when using WebFAQ as a training source?

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