Feature Aggregation from Multiple mBERT Layers for Consistent Zero-Shot Cross-Lingual Transfer

Multilingual BERT (mBERT), a language model pre-trained on large multilingual corpora, has impressive zero-shot cross-lingual transfer capabilities and performs surprisingly well on zero-shot POS tagging and Named Entity Recognition (NER), as well as on cross-lingual model transfer. At present, the mainstream methods to solve the cross-lingual downstream tasks are always using the last transformer layer's output of mBERT as the representation of linguistic information. In this work, we explore the complementary property of lower layers to the last transformer layer of mBERT. A feature aggregat

Research goal: How does aggregating features from multiple transformer layers of mBERT improve F1 score consistency in zero-shot cross-lingual transfer across XTREME-R benchmark domains compared to using only the last layer?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 8.7/10.