The TUDOK Operator - A Model-Agnostic Calibration Measure for Assessing Truth-Fidelity and Reality-Alignment in AI Outputs for Decision-Support Contexts

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

This record contains the Hungarian public pre-validation paper of the TUDOK Operator, a model-agnostic calibration measure for evaluating the truth-fidelity and reality-alignment of artificial-intelligence outputs in decision-support contexts. The paper introduces the conceptual basis, formal definition, C/K/I/D scoring logic, architectural interpretation, black-box audit applicability, diagnostic flags, pilot-ready validation protocol, scalability considerations, ethical and transparency boundaries, and open validation framework of the TUDOK Operator. The TUDOK Operator is not an objective truth-verification engine, moral authority, compliance certification, or replacement for human/domain-expert review. It is a calibration and audit-preparation instrument designed to make the truth-fidelity, uncertainty handling, source-awareness, distortion pressure, and reality-alignment behavior of AI-assisted outputs more visible, comparable, and reviewable. The open validation repository connected to this paper is available at: https://github.com/grm808-cos/tudok-operator