English(EN) A Plausible Explanation Is Enough to Mislead Multimodal Judges

合理的解释误导多模态模型中的AI裁判

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-19 04:00

研究人员开发了一个名为JUPE的新基准，用于测试多模态大型语言模型（MLLMs）对误导性解释的易感性。在对14个不同的MLLM裁判进行的实验中，合理但错误的解释显著影响了模型的判断，导致错误答案的接受率大幅提高。虽然思维链（Chain-of-Thought）推理显示出一定的缓解作用，但研究发现解释会夸大评估管道中的性能估计，而目前的防御措施保护作用有限。 AI

影响突显了AI裁判中的一个漏洞，该漏洞可能夸大性能指标，从而影响模型评估和开发。

排序理由学术论文，详细介绍了新的基准和关于AI模型行为的实验结果。[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: 12 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 学术论文，详细介绍了新的基准和关于AI模型行为的实验结果。[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: 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.AI TIER_1 English(EN) · Anglin Cai, Zhiqi Wang, Xiaozhi Yang, Yurun Lin, Dongwon Lee, Yuchen Yang · 2026-10-19 04:00

一个貌似合理的解释足以误导多模态裁判

arXiv:2610.22631v1 Announce Type: cross Abstract: Multimodal large language models (MLLMs) increasingly serve as judges in training and evaluation pipelines, often assessing answers together with explanations. Although AI judges are sensitive to textual factors, it remains unclea…

本文如何被排名

报道来源 [1]

一个貌似合理的解释足以误导多模态裁判

相关实体

相关话题