Intermediate-Task Training Effects on Zero-Shot Cross-Lingual Reasoning in 70B-Parameter LLMs

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 intermediate-task training on English language-understanding tasks (e.g., SQuAD, RACE) improve the reasoning capabilities (measured by logical consistency and accuracy on MMLU or GSM8K) of LLMs in zero-shot cross-lingual settings when scaled to 70B parameters?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 9.0/10.