Cross-Lingual Representation Alignment by Token-Level Optimal Transport in a Language-Agnostic Space
Cross-lingual alignment (CLA) aims to align the representations of large language models (LLMs) across languages, enabling cross-lingual transfer to improve multilingual capabilities. Previous CLA methods often ignore language-specific information encoded in representations and only consider sentence-level alignment, w...
Taisei Yamamoto, Ryoma Kumon, D. Bollegala et al.
· 0 citations