---
格式版本: 2
标题: "TractorBeam: Personalized AI Sensemaking Support via Collaborative Machine Annotation"
原文链接: "https://arxiv.org/abs/2608.18994"
发布日期: "2026-08-19"
发布时间校准状态: "found"
发布时间需复核: "否"
发布时间来源: "rule:local:strict_original_body"
发布时间证据: "**\\[v1\\]** Wed, 19 Aug 2026 14:59:49 UTC (1,778 KB)"
发布时间校准原因: "规则确认唯一严格发布时间，来源 local:strict_original_body"
发布时间校准置信度: "high"
发布时间候选数量: 18
发布时间严格候选数量: 6
发布时间原页读取状态: "source template page reused from URL open"
发布时间未找到原因: ""
发布时间校准时间: "2026-08-20T16:01:29+08:00"
发布时间仲裁状态: "skipped"
发布时间仲裁尝试次数: 0
发布时间仲裁耗时毫秒: 0
发现时间: "2026-08-20T15:56:53+08:00"
入库时间: "2026-08-20T08:01:29.173Z"
来源平台: "arXiv 学术论文搜索"
搜索渠道: "source_template"
搜索词: "https://arxiv.org/search/?query=AI&searchtype=all"
匹配关键词:
  - "AI"
相关厂家:
  []
相关专家:
  []
内容类型: "网页"
抓取工具: "Free Fetch + Defuddle"
清洗工具: "Defuddle Markdown + Defuddle/Readability 正文提取"
原始附件:
  []
AI优质: "否"
AI打分: 5
AI分档: "非优质"
AI质检状态: "不通过"
AI打分理由: "该资料为一篇人机交互领域的arXiv学术论文，主题是文档协作注释与AI感知支持，与超节点、AI Rack、机柜级AI基础设施等完全无关。"
AI质检模型: "ali-deepseek-v4-flash"
AI质检时间: "2026-08-20T16:08:20+08:00"
AI主题相关性: 0
AI来源权威性: 3
AI新颖性: 1
AI技术细节: 1
AI商业部署信号: 0
AI完整性: 0
AI摘要: "TractorBeam 是一个浏览器扩展式的混合主动系统，将语言模型输出重新框定为 PDF 文档上的建议高亮，通过“协作式机器标注”帮助用户在原文语境中迭代构建查询与心智模式，以改善事实性和溯源问题。"
AI摘要模型: "ali-deepseek-v4-flash"
AI摘要时间: "2026-09-07T03:19:46.768Z"
采集批次: "2026年8月20日14点19分32秒"
采集批次ID: "20260820-141932-079"
去重键: "https://arxiv.org/abs/2608.18994"
---

## Computer Science > Human-Computer Interaction

## Title:TractorBeam: Personalized AI Sensemaking Support via Collaborative Machine Annotation

Authors:[Sireesh Gururaja](https://arxiv.org/search/cs?searchtype=author&query=Gururaja,+S), [Jordan Taylor](https://arxiv.org/search/cs?searchtype=author&query=Taylor,+J), [Emma Strubell](https://arxiv.org/search/cs?searchtype=author&query=Strubell,+E)

[View PDF](https://arxiv.org/pdf/2608.18994) [HTML (experimental)](https://arxiv.org/html/2608.18994v1)

> Abstract:Language model-based systems which allow asking questions of documents have become popular tools for sensemaking. Despite their implied capability, these systems still suffer from issues of factuality and provenance, while encouraging confirmatory, rather than exploratory, research. We present TractorBeam, a browser extension-based mixed-initiative system that uses collaborative annotation as an interface metaphor for sensemaking, re-framing language model (LM) outputs as suggested highlights in a process that we call \\textit{collaborative machine annotation}. This metaphor allows us to present LM results in-context on PDF documents, directly addressing concerns of provenance and factuality, while allowing users to iteratively construct mental schemas and queries for language models directly in the context of a document. In a preliminary user study, all of our participants felt that TractorBeam enabled them evaluate and iteratively improve the model's reflection of their intended highlighting, and several found suggestions that made them reconsider their original schema. TractorBeam suggests that systems that facilitate exploratory research on individual documents may lead to verifiable sensemaking for users and complement tools that work across broader corpora.

| Comments: |  |
| --- | --- |
| Subjects: | Human-Computer Interaction (cs.HC) |
| Cite as: | [arXiv:2608.18994](https://arxiv.org/abs/2608.18994) \[cs.HC\] |
|  | (or [arXiv:2608.18994v1](https://arxiv.org/abs/2608.18994v1) \[cs.HC\] for this version) |
|  | [https://doi.org/10.48550/arXiv.2608.18994](https://doi.org/10.48550/arXiv.2608.18994) |
| Related DOI: | [https://doi.org/10.1145/3830397.3841814](https://doi.org/10.1145/3830397.3841814) |

## Submission history

From: Sireesh Gururaja \[[view email](https://arxiv.org/show-email/25e02174/2608.18994)\]  
**\[v1\]** Wed, 19 Aug 2026 14:59:49 UTC (1,778 KB)

[Which authors of this paper are endorsers?](https://arxiv.org/auth/show-endorsers/2608.18994) | Disable MathJax ([What is MathJax?](https://info.arxiv.org/help/mathjax.html))