English(EN) Signals Are Present, Predictions Fail: Probing Diagram Representations in Large Vision-Language Models

大型视觉语言模型在内部信号存在的情况下仍难以处理图关系

作者 PulseAugur 编辑部 · [1 个来源] · 2026-11-09 05:00

研究人员发现，在处理图表时，大型视觉语言模型（LVLMs）的内部表示与其预测能力之间存在脱节。尽管 LVLMs 在图表理解基准测试中表现出熟练度，但它们常常无法掌握关系，特别是涉及定向边的关系。通过对合成图表和流程图的实验，研究发现与节点和边相关的信号可以从模型的隐藏状态中解码，并且在某些情况下，这些信号会因果性地影响模型的预测。这表明模型内部处理的信息与其生成的输出之间存在差距。 AI

影响揭示了当前 LVLM 图表理解的局限性，为未来模型开发指明了方向。

排序理由学术论文，详细介绍了关于 LVLM 能力的研究发现。[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: 15 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 学术论文，详细介绍了关于 LVLM 能力的研究发现。[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.CL TIER_1 English(EN) · Haruto Yoshida, Keito Kudo, Yoichi Aoki, Ryota Tanaka, Itsumi Saito, Keisuke Sakaguchi, Kentaro Inui · 2026-11-09 05:00

信号已显现，预测却失败：探究大型视觉语言模型中的图表示

arXiv:2603.02865v2 Announce Type: replace Abstract: Large vision-language models (LVLMs) demonstrate strong performance on diagram understanding benchmarks, yet they still struggle to understand relationships between elements, particularly those represented by directed edges betw…

本文如何被排名

报道来源 [1]

信号已显现，预测却失败：探究大型视觉语言模型中的图表示

相关实体

相关话题