跳到正文
arXiv cs.MA· Eduardo Ari\~no de la Rubia, Szilard Pafka·· 9 小时前AI 评分38

Dataset Factory:智能体驱动的数据集工厂如何实现从发现到公开发布

Operating an Agent-Driven Dataset Factory: Discovery, Construction, and Public Release

AI 导读

研究团队部署 Dataset Factory,在 81.53 小时内按固定发布策略持续产出并公开发布 14 个数据集,单次发布经常性运营支出约 \$2.74,涵盖失败尝试、推理、worker 与托管成本。

正文 · 原文

View PDF HTML (experimental)

Abstract:Turning public data into prediction datasets requires connecting source discovery, task definition, construction, and distribution. We deployed Dataset Factory to perform these steps continuously under a standing release policy. Bounded language-model tasks interpret sources, write construction code, and review task meaning; deterministic software manages state, workers, retries, and publication. Over 81.53 hours, the factory published 14 datasets at an estimated recurring operating expenditure of about \$2.74 per publication. The accounting includes failed attempts, inference, workers, and hosting. Separate agents completed prediction experiments on all twelve evaluated releases. Source review nevertheless identified a task whose construction filters contradicted its population declaration. The original reviewer accepted later-record exclusions without reconciling them with the published selection rule. This experience documents an operating source-to-release workflow and shows why successful execution and consumer use need separate checks of task meaning.
Comments: 5 pages, 3 figures. Code and evidence: this https URL
Subjects: Software Engineering (cs.SE); Multiagent Systems (cs.MA)
Cite as: arXiv:2611.21957 [cs.SE]
  (or arXiv:2611.21957v1 [cs.SE] for this version)
  https://doi.org/10.48550/arXiv.2611.21957

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Eduardo Ariño De La Rubia [view email]
[v1] Mon, 16 Nov 2026 22:39:18 UTC (150 KB)

来源:arXiv cs.MA · arxiv.org