共享疑虑：语言模型的零样本跨语言置信度估计
Shared Doubt: Zero-Shot Cross-Lingual Confidence Estimation for Language Models
AI总结：
研究多语言大语言模型是否编码共享的、可跨语言迁移的置信度特征，通过轻量级线性探针从中间表示直接预测答案正确性，实现零样本跨语言泛化，并发现置信度特征集中在中间层。
摘要：
置信度估计（CE），即量化模型预测的可靠性，在大语言模型（LLM）背景下引起了极大兴趣。然而，大多数研究集中在英语上，忽视了LLM使用的多语言现实，而许多CE方法会退化或需要跨语言重新训练。为了解决这一差距，我们研究了多语言LLM是否编码共享的、可跨语言迁移的置信度特征。我们使用一个轻量级线性探针，直接从中间表示预测答案正确性。经过单语言训练后，该探针在零样本情况下泛化到未见过的、类型多样的语言，无需目标语言监督。学习到的层权重和多次消融实验表明，置信度特征集中在各语言的中间层，表明存在共享的置信度子空间。虽然零样本跨语言性能取决于与源语言的相似性，但该探针无需任何重新训练即可提供强基线，并且与其他流行的置信度估计方法相比具有优势。
英文摘要：
Confidence estimation (CE), i.e. quantifying the reliability of a model's prediction, has attracted great interest in the context of large language models (LLMs). However, most studies focus on English, ignoring the multilingual reality of LLM usage, while many CE methods degrade or require retraining across languages. To address this gap, we investigate whether multilingual LLMs encode shared, language-transferable confidence features in open-ended question answering. We use a lightweight linear probe that predicts answer correctness directly from intermediate representations. Trained monolingually, the probe generalizes zero-shot to unseen, linguistically diverse languages without target-language supervision. Multiple ablations and learned layer weights reveal that confidence features concentrate in middle layers across languages, suggesting a shared confidence subspace. While zero-shot cross-lingual performance depends on similarity to the source language, the probe provides a strong baseline without any retraining and compares favorably to other popular confidence estimation methods.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
