结构理论中的确定性：通过闭包、可比性和联合可接受性的统一框架
原标题：Determinization in Structure Theories: A Unified Framework via Closure, Comparability, and Joint Admissibility
AI 摘要
本文提出一个形式化框架，用于从多元结构理论构建规范解释，并区分了三种强度的“规范”概念：闭包稳定、全局完成和AC-6意义下的确定性。该框架将非确定性分为结构多元性（S型）和认知多元性（E型），并分别以Wyckoff和ICT理论为例。研究表明，基于完成算子的规范化需要闭包、可比性和兼容扩展条件，而基于选择器的规范化需要理论内在可比性和全局兼容性。在LLM辅助推理中，幻觉被视为无根据的规范化，即系统在底层可接受集未坍缩为单例时输出确定答案。
正文节选
Determinization in Structure Theories: A Unified Framework via Closure, Comparability, and Joint Admissibility Abstract Motivated by structural failures in LLM-assisted reasoning and decision systems, we develop a formal framework for constructing canonical interpretations from plural structure theories. A structure theory is a triple of a signature, axioms, and inference policy. Its admissible interpretation family collects all globally consistent assignments of structural conclusions to positi