新框架在回答前检测LLM中的输入歧义

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-28 04:00

研究人员开发了一个新的框架，用于在大型语言模型（LLM）尝试回答之前检测其输入的歧义。该方法使用保留标签的扰动来识别提示何时具有多种有效解释，从而发出信号表明LLM不应直接回答。在各种问答数据集上的实验表明，这种方法优于依赖LLM生成的澄清或输出端不确定性的现有方法。 AI

影响这项研究可以通过确保模型适当处理不确定或模棱两可的输入，从而实现更健壮和可靠的LLM行为。

排序理由该集群包含一篇详细介绍LLM研究新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 14 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该集群包含一篇详细介绍LLM研究新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, safety; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

arXiv cs.CL TIER_1 English(EN) · Tiejin Chen, Xiaoou Liu, Michael Clifford, Ahmadreza Moradipari, Yitao Chen, Kyungtae Han, Nejib Ammar, Hua Wei · 2026-10-28 04:00

在消歧之前进行检测：基于扰动的LLM输入歧义检测

arXiv:2610.36353v1 Announce Type: new Abstract: Large language models (LLMs) often answer ambiguous inputs as if they had a single intended meaning. This is problematic because input ambiguity is an upstream source of uncertainty: when a prompt admits multiple valid interpretatio…