analiza aiqbench
This research demonstrates that native-language sentence embeddings can effectively replace translation-based workflows in Linguistic-Integrated Reliability Auditing (LiRA) without compromising scoring accuracy or excluding valid responses due to language barriers. By validating this approach across multiple models and items, the study offers a practical pathway for scaling multilingual assessment systems while maintaining rigorous quality control standards inherent to AI-driven evaluation frameworks.
Multilingual Sentence Embeddings for Linguistic-Integrated Reliability Audit
arXiv:2607.17466v1 Announce Type: cross Abstract: Multilingual assessment systems commonly rely on translation for scoring and quality-control processes. We evaluate whether multilingual sentence embeddings can replace translated English input for Linguistic-Integrated Reliability Auditing (LiRA) across 11 PIRLS constructed-response items and three embedding models. Native-language embeddings reproduced translation-based reliability estimates closely while recovering responses excluded after translation failure, with no meaningful change in reliability.
Abstract:Multilingual assessment systems commonly rely on translation for scoring and quality-control processes. We evaluate whether multilingual sentence embeddings can replace translated English input for Linguistic-Integrated Reliability Auditing (LiRA) across 11 PIRLS constructed-response items and three embedding models. Native-language embeddings reproduced translation-based reliability estimates closely while recovering responses excluded after translation failure, with no meaningful change in reliability.
Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
arXiv:2607.17466 [cs.CL]
(or arXiv:2607.17466v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2607.17466 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 155 cuvinte
Subiecte conexe
J++ Lens: Jacobian Filtering Enables More Faithful Workspace Lenses
LessWrong · 8 oct. 2026
Continuous Language Diffusion as a Decoder-Interface Problem
arXiv · 17 iun. 2026
[AINews] Meta Muse Wonder & Wonder Full: MSL’s Claude 5.5-tier Frontier Model
Latent Space · 13 oct. 2026
Exploring Adversarial Robustness and Safety Alignment in Multilingual Multi-Modal Large Language Models
arXiv · 3 iun. 2026