Different Language
Research on different languages in the context of large language models (LLMs) focuses on understanding and mitigating biases, improving cross-lingual transfer, and enhancing model performance across diverse linguistic structures. Current efforts utilize multilingual LLMs, often employing techniques like multilingual word embeddings and decomposed prompting to improve translation and other cross-lingual tasks, while also addressing issues of data imbalance and linguistic discrimination. This work is crucial for broadening the accessibility and fairness of LLMs, impacting fields like machine translation, cross-lingual information retrieval, and language learning applications.
Papers
Multilingual Word Embeddings for Low-Resource Languages using Anchors and a Chain of Related Languages
Viktor Hangya, Silvia Severini, Radoslav Ralev, Alexander Fraser, Hinrich Schütze
PhayaThaiBERT: Enhancing a Pretrained Thai Language Model with Unassimilated Loanwords
Panyut Sriwirote, Jalinee Thapiang, Vasan Timtong, Attapol T. Rutherford