CrossIn: An Efficient Instruction Tuning Approach for Cross-Lingual Knowledge Alignment
This addresses the problem of language bias in LLMs for users needing multilingual support, but it is incremental as it builds on existing instruction tuning methods.
The paper tackles the challenge of multilingual proficiency in large language models, which often underperform in non-English languages due to imbalanced training data, by proposing CrossIn, a cross-lingual instruction tuning approach that improves performance across tasks and languages.
Multilingual proficiency presents a significant challenge for large language models (LLMs). English-centric models are usually suboptimal in other languages, particularly those that are linguistically distant from English. This performance discrepancy mainly stems from the imbalanced distribution of training data across languages during pre-training and instruction tuning stages. To address this problem, we propose a novel approach called CrossIn, which utilizes a mixed composition of cross-lingual instruction tuning data. Our method leverages the compressed representation shared by various languages to efficiently enhance the model's task-solving capabilities and multilingual proficiency within a single process. In addition, we introduce a multi-task and multi-faceted benchmark to evaluate the effectiveness of CrossIn. Experimental results demonstrate that our method substantially improves performance across tasks and languages, and we provide extensive insights into the impact of cross-lingual data volume and the integration of translation data on enhancing multilingual consistency and accuracy.