Methods for Formal Verification of Agent Skills: Three Layers Toward a Mechanically Checkable Capability-Containment Proof 文章

ArXiv CS.AI2026-05-26NEWSen作者: Alfredo Metere
查看原文 →
关系图谱

Methods for Formal Verification of Agent Skills: Three Layers Toward a Mechanically Checkable Capability-Containment Proof · 相关技术

相关技术

ORMdivide-and-conquer partitioningStraight-Through EstimatorSEMReferring expression comprehension (REC)Parts-of-Speech (POS) tagsMECLMMHISGranular Alignment ParadigmForEffort Metric AttentionEFICISAgent SkillsANNAgentICESkill