面向自适应系统的可复用LLM计划与概率保证
原标题：Reusable LLM-Based Plans with Probabilistic Guarantees for Self-Adaptive Systems
AI 摘要
该论文提出一种验证器引导的框架，将LLM编写的控制器与概率模型检查结合，为自适应系统在环境族（如不同建筑地图、不同请求数的火星车资源仲裁）中提供逐实例的概率需求保证。方法上定义了广义规则列表作为可复用符号控制器的形式化表示，并利用验证反馈让LLM生成与修复控制器。在网格世界、Pac-Man和火星探测车三个基准上评估三个LLM，LLM规则列表在约90%的留出实例上通过验证（基线最高77%），且验证已存规则列表比逐实例合成快15至103倍。
正文节选
Reusable LLM-Based Plans with Probabilistic Guarantees for Self-Adaptive Systems Abstract Self-adaptive systems are designed to accommodate variants of their environment, which form a family of instances whose uncertainty is captured by probabilistic models, and must meet probabilistic requirements on each instance. Within such a family, changes in size and structure complicate the fundamental adaptation decision: whether to retain, reuse, or construct a new controller, since controllers are oft