Skip to content
Hot storiesContinuously updated

研究提出基于嵌入语义熵的LLM不确定性估计方法

1 reports1 reporting sourcesUpdated 14 hours ago

Start here

AI overview

2026年11月4日,arXiv机器学习领域公布的一项最新研究提出了名为Embedding Semantic Entropy(Emb-SE)的新方法。该研究旨在用于大语言模型(LLM)的不确定性估计,通过在语义嵌入空间中采用K-Medoids聚类与动态选择机制,替代了传统的自然语言推理(NLI)蕴含聚类,实现了基于几何感知聚类的免NLI语义不确定性估计。

AI-generated from reports · Updated 13 hours ago

Report timeline

Follow the reports to see different sides of the story.

Nov 4
  1. arXiv 机器学习
    Embedding Semantic Entropy:基于几何感知聚类的免 NLI 语义不确定性估计

    研究提出 Embedding Semantic Entropy(Emb-SE),在语义嵌入空间采用 K-Medoids 聚类与动态选择替代传统 NLI 蕴含聚类,用于大语言模型不确定性估计。

Story heat trend

Current heat 7·Comparable peak 10(Nov 4 14:00)·Comparable change, last 24 hours–

02.557.510Nov4Nov4Nov4Nov5

Trends compare only the same parties observed continuously, so the scope can be narrower than current heat counts. Move the pointer or tap the chart to see hourly heat; use the left and right arrow keys on a keyboard.