PulseAugur
LIVE 15:53:00
tool · [1 source] ·
2
tool

LLMs struggle with Luxembourgish borrowings without knowledge graph context

Researchers have developed a new benchmark, LexNeo-Bench, to evaluate how well large language models understand lexical borrowing in low-resource languages like Luxembourgish. The benchmark, derived from a Luxembourgish news corpus, tests models on classifying borrowings and detecting neologisms. When provided with a linguistic knowledge graph as context, the models' accuracy in identifying borrowed words significantly improved, demonstrating the value of lexicon-aware prompting for evaluating LLMs in multilingual settings. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT New benchmark highlights LLM limitations in understanding nuanced linguistic phenomena like lexical borrowing in low-resource languages.

RANK_REASON Academic paper introducing a new benchmark and evaluation methodology for LLMs. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CL →

COVERAGE [1]

  1. arXiv cs.CL TIER_1 · Nina Hosseini-Kivanani ·

    Do LLMs Know What Luxembourgish Borrows? Probing Lexical Neology in Low-Resource Multilingual Models

    Large language models (LLMs) are increasingly used for writing assistance in small contact languages, yet it is unclear whether they respect community norms around lexical borrowing and neology. We introduce LexNeo-Bench, a 3{,}050-instance token-level benchmark derived from LuxB…