arXiv：cs.CL· Lihu Chen·· 4 小时前AI 评分44
大语言模型中的知识可及性几何结构
The Geometry of Knowledge Accessibility in Large Language Models
AI 导读
研究发现，LLM 的知识可及性在查询表示空间中呈现简单几何结构：越易访问的查询越靠近表示空间中心，越难访问的查询越远离中心，由此形成一条知识边界。该距离排序可跨数据集迁移，且与推理难度的关联弱于与知识可及性的关联。不同干预手段效果各异：查询改写对易访问查询更有效，思维链在边界附近更有帮助，检索则在边界之外收益更大。
正文
Abstract:Large language models (LLMs) contain broad knowledge, but they cannot access all of it reliably. We study this problem through knowledge accessibility, which describes whether the knowledge needed for a query can be recalled from the model. We find that knowledge accessibility has a simple geometric structure in the model's representation of the query alone, before any generation. More accessible queries are closer to a center in the representation space, while less accessible queries are farther away. This geometry reveals a knowledge boundary that separates more accessible queries from less accessible ones. Accessibility consistently decreases with distance from the center, and this distance-based ordering transfers across datasets even when the centers differ. Controlled experiments further show that the centered geometry is more closely related to knowledge accessibility than to reasoning difficulty. The geometry also reveals when different interventions are useful. Query rewriting helps more for accessible queries, chain-of-thought reasoning helps more near the boundary, and retrieval gives larger gains beyond the boundary. These findings not only provide a new geometric view of how knowledge is organized in language models, but also suggest a useful pre-generation signal for adaptive inference.
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.03052 [cs.CL] |
| (or arXiv:2610.03052v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.03052 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Lihu Chen [view email]
[v1]
Fri, 2 Oct 2026 09:33:10 UTC (782 KB)
来源：arXiv：cs.CL · arxiv.org