研究人员开发了一种使用大型语言模型（LLMs）通过上下文学习（ICL）填补缺失公众意见数据的新方法。该方法在调查数据上进行了测试，与MICE PMM等传统统计方法相比，显示出持续的误差减少。表现最佳的ICL方法，使用具有100个示例的gpt-oss-120b模型，实现了更窄的置信区间和更高的总体覆盖率，尤其是在非随机缺失的情况下。 AI
影响 这项研究展示了LLMs在提高公众意见数据填补的准确性和效率方面的新应用，可能影响调查方法和分析。
排序理由 学术论文，详细介绍了LLMs在数据填补方面的新应用。
在 Hugging Face Daily Papers 阅读 →
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
Large language models have been widely evaluated as simulators of individual survey responses. In practice, however, fully unobserved responses are rare; the dominant problem is partial non-response. Imputation aims to restore the overall structure of a survey dataset by filling …
Large language models have been widely evaluated as simulators of individual survey responses. In practice, however, fully unobserved responses are rare; the dominant problem is partial non-response. Imputation aims to restore the overall structure of a survey dataset by filling …