Scaling Student Model Size in Optimal Transport Distillation for Cross-Lingual Retrieval 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: Does scaling the student model size in optimal transport distillation frameworks yield diminishing returns for cross-lingual retrieval performance on low-resource languages in XTREME-R?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 9.2/10.