Cross-lingual Alignment Robustness in Intermediate-Task Training vs. Direct Tuning

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: Does English intermediate-task training degrade zero-shot cross-lingual alignment robustness on typologically distant languages within the XTREME suite compared to direct target-language tuning?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 9.3/10.