---
id: 20260313-T0-28
title: "通过不精确概率表达LLM的高阶不确定性"
title_en: "Verbalizing LLM Higher-Order Uncertainty via Imprecise Probabilities"
url: https://ai.daily.yangsir.net/daily/20260313-T0-28
issue_date: 2026-03-13
publish_date: 2026-03-12T04:00:00.000Z
source_name: "arXiv cs.AI"
source_url: https://arxiv.org/abs/2603.10396
---

# 通过不精确概率表达LLM的高阶不确定性

研究提出通过不精确概率表达LLM的高阶不确定性方法。该技术能更准确捕捉模型的置信度区间，在开放域问答任务中，不确定性估计的校准误差降低30%。

## English Version

**Verbalizing LLM Higher-Order Uncertainty via Imprecise Probabilities**

Researchers developed a method to express LLM higher-order uncertainty via imprecise probabilities. This technique better captures confidence intervals and reduces calibration error by 30% in open-domain QA tasks.

---

**来源**：[arXiv cs.AI](https://arxiv.org/abs/2603.10396)

**详情页**：https://ai.daily.yangsir.net/daily/20260313-T0-28

---

*智语观潮 · Daily — https://ai.daily.yangsir.net/llms.txt*