A research paper examines syntactic dependency trees drawn from 124 languages chosen for typological, genetic and geographic diversity. The author investigates the structural properties of these trees, noting that the organizing principles behind their topology are still not well understood. The work aims to identify regularities that hold across languages.
A new arXiv paper investigates whether traces of a writer's native language remain detectable in text after LLM-based writing tools revise it toward mainstream English conventions. Because these models are trained largely on dominant English usage, their edits may flatten the stylistic and syntactic markers that distinguish writers from different language backgrounds. The authors assess how resilient those signals are, with implications for authorship analysis and linguistic diversity.
A new arXiv paper examines word order preferences in decoder-only language models, testing 192 artificial languages alongside typologically diverse natural languages. The authors report a consistent left-branching bias in the models and argue that training data, rather than linguistic typology, shapes these preferences. The study also explores how this bias relates to model performance on right-branching languages.
Researchers built a minimal-pair dataset that tests whether language models can tell light-verb constructions like 'make a decision' apart from full predicate uses of the same verbs, such as 'make a cake'. The work targets phraseological competence, examining how verb meaning shifts in multiword expressions. By comparing model behavior on nearly identical sentence pairs, the benchmark aims to reveal how deeply models represent this distinction.
A new arXiv preprint addresses how to quantify the differences between two speakers' accents, a task relevant to both linguistics and speech technology. The authors note that measurement approaches currently differ depending on the research field, such as phonetics, and propose a method intended to be both adaptable and interpretable. The announcement provides only the abstract, so experimental details and results are not yet visible from the listing.