English(EN) Two Axes of LLM Abstention: Answer Correctness and Question Answerability

大型语言模型难以区分错误答案和无法回答的问题

作者 PulseAugur 编辑部 · [3 个来源] · 2026-07-09 13:17

研究人员确定了大型语言模型中两种不同的弃答维度：答案正确性和问题可回答性。当前模型难以区分拒绝回答不正确的问题和拒绝回答根本不应回答的问题（例如，前提错误的提问）。一种使用隐藏状态探测而非简单置信度分数的新方法，可以更好地识别无法回答的问题，从而提高模型响应的准确性。 AI

影响这项研究可能促使大型语言模型在拒绝回答不当或不正确的问题方面更加可靠，从而提高用户信任度和安全性。

排序理由该集群包含两篇在 arXiv 上发表的关于大型语言模型行为和评估的学术论文。

AI 生成摘要 · Google Gemini · 来自 3 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 0 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Research; 该集群包含两篇在 arXiv 上发表的关于大型语言模型行为和评估的学术论文。
Source corroboration: 3 independent sources; Multiple independent publishers reporting the same story raises confidence that it's real and newsworthy.
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: 91 days old; Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准。

arXiv cs.AI TIER_1 English(EN) · Benedikt J. Wagner · 2026-07-10 04:00

LLM 两种消极回答的维度：答案正确性与问题可回答性

arXiv:2607.08456v1 Announce Type: cross Abstract: A model should refuse two different things: answers it would get wrong, and questions it should not answer at all, such as unanswerable ones or ones resting on a false premise. The usual recipe thresholds a single confidence score…
arXiv cs.AI TIER_1 English(EN) · Robin Staab, Jasper Dekoninck, Maximilian Baader, Martin Vechev · 2026-07-10 04:00

为大型语言模型自适应生成引发偏见的提问

arXiv:2510.12857v2 Announce Type: replace-cross Abstract: Large language models (LLMs) are now widely deployed in user-facing applications, reaching hundreds of millions of users worldwide. Despite their widespread adoption, growing reliance on their outputs raises significant co…
arXiv cs.AI TIER_1 English(EN) · Benedikt J. Wagner · 2026-07-09 13:17

LLM 两种弃答维度：答案正确性与问题可回答性

A model should refuse two different things: answers it would get wrong, and questions it should not answer at all, such as unanswerable ones or ones resting on a false premise. The usual recipe thresholds a single confidence score, which cannot tell these apart. Across five instr…