English(EN) Each Insufficient in Its Own Way: Calibrating Language Models Across Varying Evidence States

新方法在不同证据状态下校准语言模型置信度

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-27 04:00

研究人员开发了一种名为充分性置信度估计（Sufficiency Confidence Estimation, SCE）的新方法，以改进语言模型在回答问题前评估其拥有证据的能力。与以往假设输入固定的方法不同，SCE 侧重于模型的置信度如何随着可用证据的变化而变化。通过对模型现有在下一个词元分布上的置信度进行微调，SCE 旨在将这种置信度与根据证据正确回答的实际概率相匹配。在临床诊断和多跳问答方面的实验表明，SCE 显著降低了 Brier 分数并提高了 AUROC，证明模型可以学会判断何时拥有足够的证据。 AI

影响通过使语言模型在提供答案前能更好地评估证据的充分性，提高了其可靠性。

排序理由介绍语言模型校准新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.CL 阅读 →

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: 10 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 介绍语言模型校准新方法的学术论文。[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, model release; 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: Same-day; Cluster formed today. Ranking reflects the current source set at time of score.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.CL TIER_1 English(EN) · Jiayi Li, Dennis L. Shung, Bradly C. Stadie · 2026-10-27 04:00

各自的不足：在不同证据状态下校准语言模型

arXiv:2610.33237v1 Announce Type: new Abstract: How can a language model know whether it has enough evidence to answer? Language models tend to answer even when the evidence they have is insufficient. Deciding whether to answer or gather more evidence requires a calibrated estima…

本文如何被排名

报道来源 [1]

各自的不足：在不同证据状态下校准语言模型

相关实体

相关话题