English(EN) Slow to See, Slow to Suppress: Understanding the Effects of Modality in Context-Memory Conflicts

视觉语言模型（VLMs）表现出偏好文本上下文而非视觉训练数据的偏见

作者 PulseAugur 编辑部 · [1 个来源] · 2026-09-02 04:00

一篇新的研究论文探讨了视觉语言模型（VLMs）如何处理其训练数据与提供的上下文之间冲突的信息。研究发现，VLMs表现出不对称的偏见，偏好基于文本的上下文信息，但在视觉实体方面依赖参数化（训练）数据。这归因于视觉信息处理时间更长，阻碍了现有知识的抑制。虽然思维链（chain-of-thought）推理未能解决此问题，但增加视觉上下文的数量确实显示出效果，凸显了在多模态和检索增强模型中实现一致行为所面临的挑战。 AI

影响凸显了多模态AI系统潜在的不一致性，表明在复杂应用中可靠信息处理面临挑战。

排序理由该集群包含一篇详细介绍AI模型行为研究发现的学术论文。[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: 29 / 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AI模型行为研究发现的学术论文。[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。

报道来源 [1]

arXiv cs.CL TIER_1 English(EN) · Athulith Paraselli, Etha Tianze Hua, Ellie Pavlick · 2026-09-02 04:00

迟钝的察觉，迟钝的压制：理解多模态在上下文记忆冲突中的影响

arXiv:2609.00293v1 Announce Type: new Abstract: We investigate how vision-language models (VLMs) handle context-memory conflicts; that is, situations in which the model is given information in context that differs from what was stored parametrically during training. We document a…

本文如何被排名

报道来源 [1]

迟钝的察觉，迟钝的压制：理解多模态在上下文记忆冲突中的影响

相关实体

相关话题