Multilingual and English-Only Intermediate-Task Fine-Tuning Effects on Zero-Shot Cross-Lingual Transfer Robustness

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: How does the combination of multilingual and English-only intermediate-task fine-tuning (e.g., XNLI + SNLI) affect the robustness of zero-shot cross-lingual transfer on XTREME-R, as measured by the accuracy drop under synthetic adversarial perturbations?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 8.5/10.