Thank You, Stingray: Multilingual Large Language Models Can Not (Yet) Disambiguate Cross-Lingual Word Senses

Multilingual large language models (LLMs) have gained prominence, but concerns arise regarding their reliability beyond English. This study addresses the gap in cross-lingual semantic evaluation by introducing a novel benchmark for cross-lingual sense disambiguation, StingrayBench1. In this paper, we demonstrate using false friends—words that are orthographically similar but have completely different meanings in two languages— as a possible approach to pinpoint the limitation of cross-lingual sense disambiguation in LLMs. We collect false friends in four language pairs, namely Indonesian-Malay, Indonesian-Tagalog, Chinese-Japanese, and English-German; and challenge LLMs to distinguish the use of them in context. In our analysis of various models, we observe they tend to be biased toward higher-resource languages. We also propose new metrics for quantifying the cross-lingual sense bias and comprehension based on our benchmark. Our work contributes to developing more diverse and inclusive language modeling, promoting fairer access for the wider multilingual community.

Citation

S. Cahyawijaya, R. Zhang, J.C.B. Cruz, H. Lovenia, E. Gilbert, H. Nomoto, A.F. Aji, "Thank You, Stingray: Multilingual Large Language Models Can Not (Yet) Disambiguate Cross-Lingual Word Senses," 2025, pp. 3228-3250.

Conference

Findings of the Association for Computational Linguistics: NAACL 2025

Keywords

47 Language, Communication and Culture, 4704 Linguistics

Source

Findings of the Association for Computational Linguistics: NAACL 2025

Publisher

Association for Computational Linguistics

DOI

10.18653/v1/2025.findings-naacl.178

Additional links

https://aclanthology.org/2025.findings-naacl.178/