English(EN) DocHop: Benchmarking Out-of-domain Multi-hop Reasoning in Information-Dense Documents

新的DocHop基准挑战多模态大语言模型在多跳文档推理方面的能力

作者 PulseAugur 编辑部 · [1 个来源] · 2026-09-03 04:00

研究人员推出了DocHop，这是一个旨在评估多模态大语言模型（MLLMs）在处理信息密集型文档时多跳推理能力的新基准。与评估图表和文本孤立理解的现有基准不同，DocHop要求模型利用文本上下文来选择、解释和聚合图表中的数据。该基准包含六个类别共2,074个示例，通过随机流水线生成，以控制推理深度和视觉密度。实验显示，人类标注者（准确率超过90%）与表现最佳的模型（62.83%）之间存在显著的性能差距，且随着推理复杂度的增加，性能会下降。 AI

影响 DocHop旨在推动MLLM在复杂文档理解方面的能力，可能带来更先进的用于研究和数据分析的AI助手。

排序理由该集群包含一篇介绍新AI模型评估基准的研究论文。[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: 22 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该集群包含一篇介绍新AI模型评估基准的研究论文。[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) · Zhuoran Yu, Le Thien Phuc Nguyen, Jaden Park, Xinyi Gu, Zexue He, Soochahn Lee, Rogerio Feris, Yong Jae Lee · 2026-09-03 04:00

DocHop：在信息密集型文档中进行域外多跳推理的基准测试

arXiv:2609.02059v1 Announce Type: new Abstract: Multimodal Large Language Models (MLLMs) have achieved strong performance on structured visual understanding tasks such as chart and document question answering. However, existing benchmarks typically evaluate these domains in isola…

本文如何被排名

报道来源 [1]

DocHop：在信息密集型文档中进行域外多跳推理的基准测试

相关实体

相关话题