Myeong Jun Jo

From Plausibility to Permissibility: A Layered Neuro-Symbolic Runtime for Reliable and Auditable AI Interpretation

2026 · Zenodo (CERN European Organization for Nuclear Research)

Recent generative AI systems can produce highly plausible interpretations of linguistic and operational inputs, yet plausibility alone does not guarantee execution admissibility. In real-world settings, a fluent or high-confidence output may still be semantically ill-typed, …