English(EN) Generalized Correctness Models: Learning Calibrated and Model-Agnostic Correctness Predictors from Historical Patterns

新模型使用历史数据而非自我评估来预测LLM的准确性

作者 PulseAugur 编辑部 · [1 个来源] · 2026-09-15 04:00

研究人员开发了广义正确性模型（GCMs），它们可以通过学习历史预测模式来预测大型语言模型（LLM）的准确性，而不是依赖LLM的自我评估。这些GCMs展示了一种可泛化且与模型无关的LLM置信度估计技能，在不同的数据集和模型家族中表现良好。研究发现，答案措辞是正确性的重要预测因素，并探索了诸如上下文内示例和事后校准等方法来提高预测准确性。 AI

影响这项研究通过改进置信度估计，有可能带来更可靠的LLM部署，这对于高风险应用至关重要。

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

在 arXiv cs.AI 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: 16 / 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正确性预测新方法的学术论文。[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: 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。

报道来源 [1]

arXiv cs.AI TIER_1 English(EN) · Hanqi Xiao, Vaidehi Patil, Hyunji Lee, Elias Stengel-Eskin, Mohit Bansal · 2026-09-15 04:00

通用正确性模型：从历史模式中学习校准的、模型无关的正确性预测器

arXiv:2509.24988v2 Announce Type: replace-cross Abstract: Generating accurate and calibrated confidence estimates is critical for deploying LLMs in high-stakes or user-facing applications, and remains an open challenge. Prior research has often framed confidence as a problem of e…

本文如何被排名

报道来源 [1]

通用正确性模型：从历史模式中学习校准的、模型无关的正确性预测器

相关实体

相关话题