English(EN) DocPrune:Efficient Document Question Answering via Background, Question, and Comprehension-aware Token Pruning

DocPrune框架通过令牌剪枝提升文档问答效率和准确性

作者 PulseAugur 编辑部 · [2 个来源] · 2026-04-24 06:51

研究人员推出DocPrune，一个旨在提高文档问答系统效率的新框架。该方法选择性地剪除不必要的令牌，例如背景元素或不相关文本，从而在无需额外训练的情况下降低计算负担。DocPrune还根据模型的理解水平动态识别最佳剪枝层。实验表明，在M3DocRAG基准测试中，吞吐量和准确性均得到显著提升。 AI

影响提高长文档理解任务的效率，可能降低文档AI的推理成本。

排序理由学术论文，介绍了一种新的高效文档问答方法。

AI 生成摘要 · Google Gemini · 来自 2 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; 学术论文，介绍了一种新的高效文档问答方法。
Source corroboration: 2 independent sources; Multiple independent publishers reporting the same story raises confidence that it's real and newsworthy.
Topics: paper, infra; 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: 171 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.CV TIER_1 English(EN) · Joonmyung Choi, Sanghyeok Lee, Jongha Kim, Sehyung Kim, Dohwan Ko, Jihyung Kil, Hyunwoo J. Kim · 2026-04-27 04:00

DocPrune：通过背景、问题和理解感知令牌修剪实现高效文档问答

arXiv:2604.22281v1 Announce Type: new Abstract: Recent advances in vision-language models have demonstrated remarkable performance across diverse multi-modal tasks, including document question answering that leverages structured visual cues from text, tables, and figures. However…
arXiv cs.CV TIER_1 English(EN) · Hyunwoo J. Kim · 2026-04-24 06:51

DocPrune：通过背景、问题和理解感知令牌修剪实现高效文档问答

Recent advances in vision-language models have demonstrated remarkable performance across diverse multi-modal tasks, including document question answering that leverages structured visual cues from text, tables, and figures. However, unlike natural images, document images contain…