Metraipolis › archive › dc_7d2d0279df1d
I play of the cello: Evaluating Constructional Bleeding in Multilingual LLMs with CALQUE — arXiv:2611.10266
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.10266
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-10T18:44:25.077Z
- SHA-256 of the archived text
bb936abcaf8a11a051e1f05cb94578ea6eba76beec64c285bf931243b04f6ced- Stable URL of this page
https://metraipolis.com/documento/dc_7d2d0279df1d
How to cite
I play of the cello: Evaluating Constructional Bleeding in Multilingual LLMs with CALQUE — arXiv:2611.10266 — arXiv (en), https://arxiv.org/abs/2611.10266 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7d2d0279df1d (Metraipolis archive dc_7d2d0279df1d, sha256 bb936abcaf8a11a051e1f05cb94578ea6eba76beec64c285bf931243b04f6ced)
BibTeX
@misc{dc7d2d0279df1d,
title = {I play of the cello: Evaluating Constructional Bleeding in Multilingual LLMs with CALQUE — arXiv:2611.10266},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.10266},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7d2d0279df1d; sha256 bb936abcaf8a11a051e1f05cb94578ea6eba76beec64c285bf931243b04f6ced},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.10266. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Languages use different morphosyntactic strategies to express the same meaning. Speakers of multiple languages are able to apply strategies from one language to another, either by mistake or intentionally, to produce a literal translation, or calque. Considering that large language models (LLMs) have been shown to generalize crosslingually, despite being predominantly trained on English, we investigate whether they overgeneralize constructional strategies from English to other languages. To test this, we create a dataset of 5583 sentence pairs across seven languages, matching literal translations of strategies from English into other languages, and literal translations of the other languages' strategies into English. We then quantify constructional bleeding by measuring how often strategies from English are assigned a higher probability than native strategies. We find that models often fail to assign higher probabilities to native strategies and tend to prefer English-derived strategies even when evaluating sentences in other languages. Our findings highlight the limitations of the current English-dominant training paradigm. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno). Categorie arXiv: Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2611.10266 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-11-10T18:37:52.421Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).