arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2105.11267cs.AIcs.PL
收藏

您可以处理的操作:AI计划的从属类型

Actions You Can Handle: Dependent Types for AI Plans

Heriot-Watt University(赫瑞-瓦特大学)

Alasdair Hill, Ekaterina Komendantskaya, Matthew L. Daggitt, Ronald P. A. Petrick

摘要:

人工智能的验证是一项挑战,涉及工程、算法和程序设计语言等方面。例如,人工智能规划者被部署来模拟自主智能体的动作。它们包括许多搜索算法,给定一组指定的属性,查找满足这些属性的动作序列。虽然从算法和工程的角度来看,人工智能规划器是成熟的工具,但它们作为程序设计语言有局限性。可判定和有效的自动搜索需要对语言的语法进行限制,禁止使用高阶属性或递归。本文提出了一种将人工智能规划器生成的计划嵌入到依赖类型语言Agda中的方法,该方法使用户能够推理和验证计划的更一般和抽象的属性,并且还为建模计划执行提供了更全面的程序设计语言基础设施。

英文摘要:

Verification of AI is a challenge that has engineering, algorithmic and programming language components. For example, AI planners are deployed to model actions of autonomous agents. They comprise a number of searching algorithms that, given a set of specified properties, find a sequence of actions that satisfy these properties. Although AI planners are mature tools from the algorithmic and engineering points of view, they have limitations as programming languages. Decidable and efficient automated search entails restrictions on the syntax of the language, prohibiting use of higher-order properties or recursion. This paper proposes a methodology for embedding plans produced by AI planners into dependently-typed language Agda, which enables users to reason about and verify more general and abstract properties of plans, and also provides a more holistic programming language infrastructure for modelling plan execution.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑