arXiv cs.CL· Abhishek Barali, Aayush Paudel·· 13 小时前AI 评分39
研究：Schema 中的最小条目数会让开源权重模型在抽取列表中凑数填充
Padding to k: A Minimum Count in the Schema Makes Open-Weight Models Pad Their Extraction Lists
AI 导读
一项预注册研究测试 8 个开源权重模型和 6 个托管服务模型，发现 schema 中未强制的最小条目数会被模型当作目标：列表超出真实条目数的比例从 34.5% 升至 55.4%。多数填充条目是文档中的真实文本，约一半能通过"条目出现在原文中"的检查。常见修复仅部分有效：提示"条目更少也可以"平均可消除该效应但会漏掉更多真实条目，而带校验错误重问会让填充更普遍。
正文 · 原文
Abstract:Extraction pipelines ask language models for JSON that follows a schema, and the schema can set a minimum number of list entries. We ask what models do when the document holds fewer items than that minimum. Under a preregistered plan, we test eight open-weight models on 558 real documents from nine annotated datasets, with six more models on hosted services as a reference. Three findings stand out. First, models treat the count as a target: a minimum shown in the schema, but not enforced, raises the share of lists longer than the true count from 34.5% to 55.4%. Second, the padding is hard to catch, because most added entries are real text taken from the document; about half of them pass a check that each entry occurs in the source. Third, common fixes help only partly: a line saying fewer items are fine removes the effect on average but misses more true items, a count in prose ("up to k") also raises over-counting, an "N/A" option under an enforced minimum still leaves half the lists padded, and re-asking with the validation error makes padding far more common. We release the code, answer lists and run records at this https URL
| Comments: | 26 pages, 4 figures. Preregistered at this https URL. Code and data: this https URL |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.30997 [cs.CL] |
| (or arXiv:2610.30997v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.30997 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Abhishek Barali [view email]
[v1]
Thu, 22 Oct 2026 09:38:58 UTC (66 KB)
来源：arXiv cs.CL · arxiv.org