arXiv cs.CL· Ryotaro Shimizu, Sai HtaungKham, Mo Fan·· 13 小时前AI 评分38
偏好不等于授权：智能体电商中的自动购买边界个性化
Preference Is Not Permission: Personalizing the Auto-Purchase Line in Agentic Commerce
AI 导读
研究提出“自动购买线”任务，让智能体判断在商品匹配度之外是否可直接下单，输出不买、先确认或自动购买三值结果。在 16 个指令微调模型上，模型普遍在人货匹配度低、自身证据标记缺失时仍自主下单，与 15 名标注者给出的授权参考不一致。基于个人数十条标注训练的浅层学习器在 30-40 条标签时中位预测达 0.79，优于所有提示词×模型配置，且该边界不跨领域迁移。
正文 · 原文
Abstract:Personalized shopping benchmarks ask which item fits the user; we ask how far an agent may commit on that item, and pose learning the individual's auto-purchase line as a task. The boundary is evidence-conditioned and three-valued: whether committing is appropriate depends on the sufficiency of product information (warrant), reversibility, price against the user's norm, and urgency, not only on fit; and the natural output is don't buy, confirm first, or buy automatically. Across sixteen instruction-tuned models, autonomous commitment is misaligned with a human authorization reference in which fifteen annotators, role-playing the delegating principal from a given profile, choose the furthest action they would allow without confirmation: models commit even when fit is low and when their own evidence flag reports the warrant missing. This is overreach. The line is individual: these raters disagree, yet the disagreement is structured and partly predictable from a person's own other labels. A shallow learner on a few dozen of a person's own labels predicts their held-out auto-purchase line at a median of 0.79 by 30-40 labels, outperforming every evaluated prompt$\times$model configuration, including a per-annotator oracle over 80 configurations; the learned boundary does not transfer across domains. We release the counterfactual benchmark, human authorization reference, and annotation protocol after acceptance.
| Subjects: | Information Retrieval (cs.IR); Computation and Language (cs.CL); Human-Computer Interaction (cs.HC); Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.29216 [cs.IR] |
| (or arXiv:2610.29216v1 [cs.IR] for this version) | |
| https://doi.org/10.48550/arXiv.2610.29216 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Ryotaro Shimizu [view email]
[v1]
Wed, 21 Oct 2026 05:15:31 UTC (112 KB)
来源：arXiv cs.CL · arxiv.org