Multilingual Large Language Model
Multilingual large language models (MLLMs) aim to extend the capabilities of large language models to multiple languages, improving cross-lingual understanding and generation. Current research focuses on enhancing performance for low-resource languages through techniques like continued pre-training on massive multilingual datasets, parameter-efficient fine-tuning with knowledge graphs, and mitigating biases and improving factual accuracy across languages. These advancements are significant for bridging the language gap in AI applications, fostering inclusivity, and enabling more equitable access to advanced language technologies globally.
Papers
Decomposed Prompting: Unveiling Multilingual Linguistic Structure Knowledge in English-Centric Large Language Models
Ercong Nie, Shuzhou Yuan, Bolei Ma, Helmut Schmid, Michael Färber, Frauke Kreuter, Hinrich Schütze
Multi-FAct: Assessing Factuality of Multilingual LLMs using FActScore
Sheikh Shafayat, Eunsu Kim, Juhyun Oh, Alice Oh
Could We Have Had Better Multilingual LLMs If English Was Not the Central Language?
Ryandito Diandaru, Lucky Susanto, Zilu Tang, Ayu Purwarianti, Derry Wijaya
OMGEval: An Open Multilingual Generative Evaluation Benchmark for Large Language Models
Yang Liu, Meng Xu, Shuo Wang, Liner Yang, Haoyu Wang, Zhenghao Liu, Cunliang Kong, Yun Chen, Yang Liu, Maosong Sun, Erhong Yang