arXiv cs.CL· Federico Marcuzzi, Martin Christoph, Iryna Gurevych, Alla Rozovskays·· 13 小时前AI 评分31
研究发现非母语者的语法错误会加剧大语言模型的社会偏见
Investigating the Impact of Learner Errors on Social Bias in Large Language Models
AI 导读
一项受控实验向三个社会偏见基准的提示词中注入常见学习者语法错误，发现语法错误会加剧模型偏见，且错误率越高偏见越强。四类错误中，介词错误对刻板印象对齐的增幅最大且最一致，分析显示错误越可能干扰理解就越容易推高偏见。在含错误文本上微调可削弱这一偏见增幅，说明仅用规范英语提示词的偏见评测可能低估非母语用户遭遇的偏见。
正文 · 原文
Abstract:Existing social bias benchmarks evaluate large language models mostly on fluent, well-formed English, reflecting native speaker usage and leaving non-native writers unrepresented. To address this gap, we investigate how grammatical errors by non-native writers affect social bias. We develop a controlled framework that injects common learner errors into prompts and apply it to three bias benchmarks covering stereotype alignment and demographic preference across gender, race, and religion. We find that grammatical errors increase bias, and the effect grows with the error rate. Among the four common error types we consider, preposition errors produce the largest and most consistent increase in stereotype alignment. Our analysis suggests that errors increase bias when they are more likely to interfere with comprehension. Finally, fine-tuning on erroneous texts reduces the bias increase caused by grammatical errors. Our findings show that bias evaluations based exclusively on well-formed prompts may underestimate the bias experienced by users whose English contains grammatical errors.
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.29896 [cs.CL] |
| (or arXiv:2610.29896v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.29896 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Federico Marcuzzi [view email]
[v1]
Wed, 21 Oct 2026 15:38:13 UTC (763 KB)
来源：arXiv cs.CL · arxiv.org