Native Japanese raters and six LLMs both penalize L2-written Japanese emails on fluency, status, and solidarity, with AIs reproducing the bias in attenuated form and over-differentiating learner backgrounds.
Through either route, these attitudes contribute to discrimination in employment, education, and legal contexts (Craft et al., 2020)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Human-LLM Alignment in Language Attitudes Toward Non-Native Japanese
Native Japanese raters and six LLMs both penalize L2-written Japanese emails on fluency, status, and solidarity, with AIs reproducing the bias in attenuated form and over-differentiating learner backgrounds.