версии (v1, preprint)
Year: 2026
Type: article
Abstract: Ontology-Preserving Mapping Theory (OPMT):A Homomorphic Framework for AI Safety and ModelauditingAndrey A. ArtsybashevIndependent Researcher, Kharkiv, UkraineIdentifier: AAM-V1_ARTSYBASHEV_UA_KHARKIV_AIANALYSISFebruary 9, 2026AbstractAs Large Language Models (LLMs) and generative AI systems become integral to Research& Development (R&D), the risk of “hallucinations” shifts from semantic incoherence to on-tological invalidityplausible but physically impossible descriptions. This paper formalizes theArtsybashev Analysis Method (AAM-V1) and the AAM-RSL v1.2 (Responsibility& Skepticism Layer) protocol. We introduce the concept of Ontological Homomorphism, astructural mapping Φ : R → M that preserves physical invariants (entropy, energy, causality)between reality (R) and the model (M ). We classify model outputs into VALID (homomor-phism preserved), FRINGE (partial homomorphism with a large kernel), and GHOST(structural violation). Using the PseudoPhysicsAI case study, we demonstrate how thisframework detects subtle violations of thermodynamic laws, providing a rigorous tool forauditing AI in high-responsibility domains.Keywords: AI Safety, Ontological Homomorphism, AAM-RSL, Hallucination Detection,R&D, Entropy, Epistemology.
Author: ARTSYBASHEV, ANDRII
Language: English
Subfield: Safety Research
Field: Social Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7128614972
FWCI:
Cites:
Cited by:
Related to: