Impact of Multilingual Intermediate-Task Training on Zero-Shot Accuracy Across Typologically Diverse Languages in XTREME-R

Intermediate-task training---fine-tuning a pretrained model on an intermediate task before fine-tuning again on the target task---often improves model performance substantially on language understanding tasks in monolingual English settings. We investigate whether English intermediate-task training is still helpful on non-English target tasks. Using nine intermediate language-understanding tasks, we evaluate intermediate-task transfer in a zero-shot cross-lingual setting on the XTREME benchmark. We see large improvements from intermediate training on the BUCC and Tatoeba sentence retrieval tas

Research goal: What is the impact of multilingual intermediate-task training (using multiple source languages) on zero-shot accuracy across typologically diverse languages in XTREME-R compared to monolingual English intermediate-task training?

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