arXiv cs.CL· Nicholas S. Kersting, Vittorio Castelli, Chieh Ting Yeh, Xinzhu Wang, Saad Taame, Khaoula Allak·· 11 小时前AI 评分31
文本语料作为概念场：提供可解释的接地性
Text Corpora as Concept Fields: Providing Explainable Groundedness
AI 导读
研究者提出“概念场”（Concept Field），在句子嵌入空间中用相邻句子的 delta 估计带逐点不确定性的局部漂移场，并以 ζ 分数衡量候选句子转移与场的吻合度，给出模型无关、可归因于语料的接地性量化指标。
正文 · 原文
Abstract:We introduce the \textbf{Concept Field} of a text corpus: a local drift field with pointwise uncertainty, estimated in sentence-embedding space from the deltas between consecutive sentences. Given a candidate sentence transition, we score its agreement with the field by $\zeta$, the mean absolute z-distance between the observed delta and the field's local statistical estimate, giving a quantitative measure of groundedness. The score is model agnostic and corpus-attributable, providing traces to corpus semantics either supporting or refuting groundedness. We drive the computation with the introduction of a \textbf{Vector Sequence Database (VSDB)} that stores embeddings together with sequence-position and next-delta metadata. We demonstrate this approach on two large-scale settings: groundedness detection over the U.S. Code of Federal Regulations, and ungroundedness (novelty) detection over Project Gutenberg, where we show Concept Fields achieve strong selective classification performance under a grounded / ungrounded / unsure triage policy. Unlike retrieval-centric baselines, the resulting coverage-risk behavior shows similar empirical coverage-performance trends in the two evaluated domains. Concept Fields provide a fast, lightweight, and interpretable signal for groundedness, complementary to LLM-as-judge and white-box detectors.
| Comments: | 33 pages, 10 figures, 20 tables; additional explainability analysis inserted |
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computers and Society (cs.CY) |
| Cite as: | arXiv:2605.05103 [cs.CL] |
| (or arXiv:2605.05103v4 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2605.05103 arXiv-issued DOI via DataCite |
Submission history
From: Nicholas Kersting [view email]
[v1]
Wed, 6 May 2026 16:38:49 UTC (1,694 KB)
[v2]
Sat, 9 May 2026 17:34:20 UTC (1,694 KB)
[v3]
Wed, 12 Aug 2026 16:10:50 UTC (1,928 KB)
[v4]
Fri, 30 Oct 2026 20:11:45 UTC (2,441 KB)
来源：arXiv cs.CL · arxiv.org